From 71d8e94dabee7fceac473d87445a03e848469a71 Mon Sep 17 00:00:00 2001 From: Moritz Fischer Date: Wed, 26 Jun 2019 17:33:09 -0700 Subject: fpga: altera-pr-ip: Make alt_pr_unregister function void Make alt_pr_unregister function void, since it always returns 0, and nothing would act on the value anyways. Signed-off-by: Moritz Fischer --- include/linux/fpga/altera-pr-ip-core.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/fpga/altera-pr-ip-core.h b/include/linux/fpga/altera-pr-ip-core.h index 7d4664730d60..0b08ac20ab16 100644 --- a/include/linux/fpga/altera-pr-ip-core.h +++ b/include/linux/fpga/altera-pr-ip-core.h @@ -13,6 +13,6 @@ #include int alt_pr_register(struct device *dev, void __iomem *reg_base); -int alt_pr_unregister(struct device *dev); +void alt_pr_unregister(struct device *dev); #endif /* _ALT_PR_IP_CORE_H */ -- cgit v1.2.3-59-g8ed1b From 615c164da0eb42cbfb1688cb429cc4d5039db5d8 Mon Sep 17 00:00:00 2001 From: Alexander Shishkin Date: Fri, 5 Jul 2019 17:14:21 +0300 Subject: intel_th: msu: Introduce buffer interface Introduces a concept of external buffers, which is a mechanism for creating trace sinks that would receive trace data from MSC buffers and transfer it elsewhere. A external buffer can implement its own window allocation/deallocation if it has to. It must provide a callback that's used to notify it when a window fills up, so that it can then start a DMA transaction from that window 'elsewhere'. This window remains in a 'locked' state and won't be used for storing new trace data until the buffer 'unlocks' it with a provided API call, at which point the window can be used again for storing trace data. This relies on a functional "last block" interrupt, so not all versions of Trace Hub can use this feature, which does not reflect on existing users. Signed-off-by: Alexander Shishkin Reviewed-by: Andy Shevchenko Link: https://lore.kernel.org/r/20190705141425.19894-2-alexander.shishkin@linux.intel.com Signed-off-by: Greg Kroah-Hartman --- .../ABI/testing/sysfs-bus-intel_th-devices-msc | 3 +- MAINTAINERS | 1 + drivers/hwtracing/intel_th/msu.c | 388 +++++++++++++++++++-- drivers/hwtracing/intel_th/msu.h | 20 +- include/linux/intel_th.h | 79 +++++ 5 files changed, 461 insertions(+), 30 deletions(-) create mode 100644 include/linux/intel_th.h (limited to 'include') diff --git a/Documentation/ABI/testing/sysfs-bus-intel_th-devices-msc b/Documentation/ABI/testing/sysfs-bus-intel_th-devices-msc index f54ae244f3f1..456cb62b384c 100644 --- a/Documentation/ABI/testing/sysfs-bus-intel_th-devices-msc +++ b/Documentation/ABI/testing/sysfs-bus-intel_th-devices-msc @@ -12,7 +12,8 @@ Description: (RW) Configure MSC operating mode: - "single", for contiguous buffer mode (high-order alloc); - "multi", for multiblock mode; - "ExI", for DCI handler mode; - - "debug", for debug mode. + - "debug", for debug mode; + - any of the currently loaded buffer sinks. If operating mode changes, existing buffer is deallocated, provided there are no active users and tracing is not enabled, otherwise the write will fail. diff --git a/MAINTAINERS b/MAINTAINERS index 783569e3c4b4..c8c506b8423b 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -8360,6 +8360,7 @@ M: Alexander Shishkin S: Supported F: Documentation/trace/intel_th.rst F: drivers/hwtracing/intel_th/ +F: include/linux/intel_th.h INTEL(R) TRUSTED EXECUTION TECHNOLOGY (TXT) M: Ning Sun diff --git a/drivers/hwtracing/intel_th/msu.c b/drivers/hwtracing/intel_th/msu.c index 8ab28e5fb366..08413e6a075f 100644 --- a/drivers/hwtracing/intel_th/msu.c +++ b/drivers/hwtracing/intel_th/msu.c @@ -17,21 +17,48 @@ #include #include #include +#include #include #ifdef CONFIG_X86 #include #endif +#include #include "intel_th.h" #include "msu.h" #define msc_dev(x) (&(x)->thdev->dev) +/* + * Lockout state transitions: + * READY -> INUSE -+-> LOCKED -+-> READY -> etc. + * \-----------/ + * WIN_READY: window can be used by HW + * WIN_INUSE: window is in use + * WIN_LOCKED: window is filled up and is being processed by the buffer + * handling code + * + * All state transitions happen automatically, except for the LOCKED->READY, + * which needs to be signalled by the buffer code by calling + * intel_th_msc_window_unlock(). + * + * When the interrupt handler has to switch to the next window, it checks + * whether it's READY, and if it is, it performs the switch and tracing + * continues. If it's LOCKED, it stops the trace. + */ +enum lockout_state { + WIN_READY = 0, + WIN_INUSE, + WIN_LOCKED +}; + /** * struct msc_window - multiblock mode window descriptor * @entry: window list linkage (msc::win_list) * @pgoff: page offset into the buffer that this window starts at + * @lockout: lockout state, see comment below + * @lo_lock: lockout state serialization * @nr_blocks: number of blocks (pages) in this window * @nr_segs: number of segments in this window (<= @nr_blocks) * @_sgt: array of block descriptors @@ -40,6 +67,8 @@ struct msc_window { struct list_head entry; unsigned long pgoff; + enum lockout_state lockout; + spinlock_t lo_lock; unsigned int nr_blocks; unsigned int nr_segs; struct msc *msc; @@ -77,6 +106,8 @@ struct msc_iter { * struct msc - MSC device representation * @reg_base: register window base address * @thdev: intel_th_device pointer + * @mbuf: MSU buffer, if assigned + * @mbuf_priv MSU buffer's private data, if @mbuf * @win_list: list of windows in multiblock mode * @single_sgt: single mode buffer * @cur_win: current window @@ -100,6 +131,10 @@ struct msc { void __iomem *msu_base; struct intel_th_device *thdev; + const struct msu_buffer *mbuf; + void *mbuf_priv; + + struct work_struct work; struct list_head win_list; struct sg_table single_sgt; struct msc_window *cur_win; @@ -126,6 +161,101 @@ struct msc { unsigned int index; }; +static LIST_HEAD(msu_buffer_list); +static struct mutex msu_buffer_mutex; + +/** + * struct msu_buffer_entry - internal MSU buffer bookkeeping + * @entry: link to msu_buffer_list + * @mbuf: MSU buffer object + * @owner: module that provides this MSU buffer + */ +struct msu_buffer_entry { + struct list_head entry; + const struct msu_buffer *mbuf; + struct module *owner; +}; + +static struct msu_buffer_entry *__msu_buffer_entry_find(const char *name) +{ + struct msu_buffer_entry *mbe; + + lockdep_assert_held(&msu_buffer_mutex); + + list_for_each_entry(mbe, &msu_buffer_list, entry) { + if (!strcmp(mbe->mbuf->name, name)) + return mbe; + } + + return NULL; +} + +static const struct msu_buffer * +msu_buffer_get(const char *name) +{ + struct msu_buffer_entry *mbe; + + mutex_lock(&msu_buffer_mutex); + mbe = __msu_buffer_entry_find(name); + if (mbe && !try_module_get(mbe->owner)) + mbe = NULL; + mutex_unlock(&msu_buffer_mutex); + + return mbe ? mbe->mbuf : NULL; +} + +static void msu_buffer_put(const struct msu_buffer *mbuf) +{ + struct msu_buffer_entry *mbe; + + mutex_lock(&msu_buffer_mutex); + mbe = __msu_buffer_entry_find(mbuf->name); + if (mbe) + module_put(mbe->owner); + mutex_unlock(&msu_buffer_mutex); +} + +int intel_th_msu_buffer_register(const struct msu_buffer *mbuf, + struct module *owner) +{ + struct msu_buffer_entry *mbe; + int ret = 0; + + mbe = kzalloc(sizeof(*mbe), GFP_KERNEL); + if (!mbe) + return -ENOMEM; + + mutex_lock(&msu_buffer_mutex); + if (__msu_buffer_entry_find(mbuf->name)) { + ret = -EEXIST; + kfree(mbe); + goto unlock; + } + + mbe->mbuf = mbuf; + mbe->owner = owner; + list_add_tail(&mbe->entry, &msu_buffer_list); +unlock: + mutex_unlock(&msu_buffer_mutex); + + return ret; +} +EXPORT_SYMBOL_GPL(intel_th_msu_buffer_register); + +void intel_th_msu_buffer_unregister(const struct msu_buffer *mbuf) +{ + struct msu_buffer_entry *mbe; + + mutex_lock(&msu_buffer_mutex); + mbe = __msu_buffer_entry_find(mbuf->name); + if (mbe) { + list_del(&mbe->entry); + kfree(mbe); + } + mutex_unlock(&msu_buffer_mutex); +} +EXPORT_SYMBOL_GPL(intel_th_msu_buffer_unregister); + static inline bool msc_block_is_empty(struct msc_block_desc *bdesc) { /* header hasn't been written */ @@ -188,6 +318,25 @@ static struct msc_window *msc_next_window(struct msc_window *win) return list_next_entry(win, entry); } +static size_t msc_win_total_sz(struct msc_window *win) +{ + unsigned int blk; + size_t size = 0; + + for (blk = 0; blk < win->nr_segs; blk++) { + struct msc_block_desc *bdesc = msc_win_block(win, blk); + + if (msc_block_wrapped(bdesc)) + return win->nr_blocks << PAGE_SHIFT; + + size += msc_total_sz(bdesc); + if (msc_block_last_written(bdesc)) + break; + } + + return size; +} + /** * msc_find_window() - find a window matching a given sg_table * @msc: MSC device @@ -527,6 +676,9 @@ static int intel_th_msu_init(struct msc *msc) if (!msc->do_irq) return 0; + if (!msc->mbuf) + return 0; + mintctl = ioread32(msc->msu_base + REG_MSU_MINTCTL); mintctl |= msc->index ? M1BLIE : M0BLIE; iowrite32(mintctl, msc->msu_base + REG_MSU_MINTCTL); @@ -554,6 +706,44 @@ static void intel_th_msu_deinit(struct msc *msc) iowrite32(mintctl, msc->msu_base + REG_MSU_MINTCTL); } +static int msc_win_set_lockout(struct msc_window *win, + enum lockout_state expect, + enum lockout_state new) +{ + enum lockout_state old; + unsigned long flags; + int ret = 0; + + if (!win->msc->mbuf) + return 0; + + spin_lock_irqsave(&win->lo_lock, flags); + old = win->lockout; + + if (old != expect) { + ret = -EINVAL; + dev_warn_ratelimited(msc_dev(win->msc), + "expected lockout state %d, got %d\n", + expect, old); + goto unlock; + } + + win->lockout = new; + +unlock: + spin_unlock_irqrestore(&win->lo_lock, flags); + + if (ret) { + if (expect == WIN_READY && old == WIN_LOCKED) + return -EBUSY; + + /* from intel_th_msc_window_unlock(), don't warn if not locked */ + if (expect == WIN_LOCKED && old == new) + return 0; + } + + return ret; +} /** * msc_configure() - set up MSC hardware * @msc: the MSC device to configure @@ -571,8 +761,12 @@ static int msc_configure(struct msc *msc) if (msc->mode > MSC_MODE_MULTI) return -ENOTSUPP; - if (msc->mode == MSC_MODE_MULTI) + if (msc->mode == MSC_MODE_MULTI) { + if (msc_win_set_lockout(msc->cur_win, WIN_READY, WIN_INUSE)) + return -EBUSY; + msc_buffer_clear_hw_header(msc); + } reg = msc->base_addr >> PAGE_SHIFT; iowrite32(reg, msc->reg_base + REG_MSU_MSC0BAR); @@ -594,10 +788,14 @@ static int msc_configure(struct msc *msc) iowrite32(reg, msc->reg_base + REG_MSU_MSC0CTL); + intel_th_msu_init(msc); + msc->thdev->output.multiblock = msc->mode == MSC_MODE_MULTI; intel_th_trace_enable(msc->thdev); msc->enabled = 1; + if (msc->mbuf && msc->mbuf->activate) + msc->mbuf->activate(msc->mbuf_priv); return 0; } @@ -611,10 +809,17 @@ static int msc_configure(struct msc *msc) */ static void msc_disable(struct msc *msc) { + struct msc_window *win = msc->cur_win; u32 reg; lockdep_assert_held(&msc->buf_mutex); + if (msc->mode == MSC_MODE_MULTI) + msc_win_set_lockout(win, WIN_INUSE, WIN_LOCKED); + + if (msc->mbuf && msc->mbuf->deactivate) + msc->mbuf->deactivate(msc->mbuf_priv); + intel_th_msu_deinit(msc); intel_th_trace_disable(msc->thdev); if (msc->mode == MSC_MODE_SINGLE) { @@ -630,6 +835,11 @@ static void msc_disable(struct msc *msc) reg = ioread32(msc->reg_base + REG_MSU_MSC0CTL); reg &= ~MSC_EN; iowrite32(reg, msc->reg_base + REG_MSU_MSC0CTL); + + if (msc->mbuf && msc->mbuf->ready) + msc->mbuf->ready(msc->mbuf_priv, win->sgt, + msc_win_total_sz(win)); + msc->enabled = 0; iowrite32(0, msc->reg_base + REG_MSU_MSC0BAR); @@ -640,6 +850,10 @@ static void msc_disable(struct msc *msc) reg = ioread32(msc->reg_base + REG_MSU_MSC0STS); dev_dbg(msc_dev(msc), "MSCnSTS: %08x\n", reg); + + reg = ioread32(msc->reg_base + REG_MSU_MSUSTS); + reg &= msc->index ? MSUSTS_MSC1BLAST : MSUSTS_MSC0BLAST; + iowrite32(reg, msc->reg_base + REG_MSU_MSUSTS); } static int intel_th_msc_activate(struct intel_th_device *thdev) @@ -856,6 +1070,8 @@ static int msc_buffer_win_alloc(struct msc *msc, unsigned int nr_blocks) win->msc = msc; win->sgt = &win->_sgt; + win->lockout = WIN_READY; + spin_lock_init(&win->lo_lock); if (!list_empty(&msc->win_list)) { struct msc_window *prev = list_last_entry(&msc->win_list, @@ -865,8 +1081,13 @@ static int msc_buffer_win_alloc(struct msc *msc, unsigned int nr_blocks) win->pgoff = prev->pgoff + prev->nr_blocks; } - ret = __msc_buffer_win_alloc(win, nr_blocks); - if (ret < 0) + if (msc->mbuf && msc->mbuf->alloc_window) + ret = msc->mbuf->alloc_window(msc->mbuf_priv, &win->sgt, + nr_blocks << PAGE_SHIFT); + else + ret = __msc_buffer_win_alloc(win, nr_blocks); + + if (ret <= 0) goto err_nomem; msc_buffer_set_uc(win, ret); @@ -925,7 +1146,10 @@ static void msc_buffer_win_free(struct msc *msc, struct msc_window *win) msc_buffer_set_wb(win); - __msc_buffer_win_free(msc, win); + if (msc->mbuf && msc->mbuf->free_window) + msc->mbuf->free_window(msc->mbuf_priv, win->sgt); + else + __msc_buffer_win_free(msc, win); kfree(win); } @@ -1462,18 +1686,77 @@ static void msc_win_switch(struct msc *msc) intel_th_trace_switch(msc->thdev); } +/** + * intel_th_msc_window_unlock - put the window back in rotation + * @dev: MSC device to which this relates + * @sgt: buffer's sg_table for the window, does nothing if NULL + */ +void intel_th_msc_window_unlock(struct device *dev, struct sg_table *sgt) +{ + struct msc *msc = dev_get_drvdata(dev); + struct msc_window *win; + + if (!sgt) + return; + + win = msc_find_window(msc, sgt, false); + if (!win) + return; + + msc_win_set_lockout(win, WIN_LOCKED, WIN_READY); +} +EXPORT_SYMBOL_GPL(intel_th_msc_window_unlock); + +static void msc_work(struct work_struct *work) +{ + struct msc *msc = container_of(work, struct msc, work); + + intel_th_msc_deactivate(msc->thdev); +} + static irqreturn_t intel_th_msc_interrupt(struct intel_th_device *thdev) { struct msc *msc = dev_get_drvdata(&thdev->dev); u32 msusts = ioread32(msc->msu_base + REG_MSU_MSUSTS); u32 mask = msc->index ? MSUSTS_MSC1BLAST : MSUSTS_MSC0BLAST; + struct msc_window *win, *next_win; - if (!(msusts & mask)) { - if (msc->enabled) - return IRQ_HANDLED; + if (!msc->do_irq || !msc->mbuf) return IRQ_NONE; + + msusts &= mask; + + if (!msusts) + return msc->enabled ? IRQ_HANDLED : IRQ_NONE; + + iowrite32(msusts, msc->msu_base + REG_MSU_MSUSTS); + + if (!msc->enabled) + return IRQ_NONE; + + /* grab the window before we do the switch */ + win = msc->cur_win; + if (!win) + return IRQ_HANDLED; + next_win = msc_next_window(win); + if (!next_win) + return IRQ_HANDLED; + + /* next window: if READY, proceed, if LOCKED, stop the trace */ + if (msc_win_set_lockout(next_win, WIN_READY, WIN_INUSE)) { + schedule_work(&msc->work); + return IRQ_HANDLED; } + /* current window: INUSE -> LOCKED */ + msc_win_set_lockout(win, WIN_INUSE, WIN_LOCKED); + + msc_win_switch(msc); + + if (msc->mbuf && msc->mbuf->ready) + msc->mbuf->ready(msc->mbuf_priv, win->sgt, + msc_win_total_sz(win)); + return IRQ_HANDLED; } @@ -1511,21 +1794,43 @@ wrap_store(struct device *dev, struct device_attribute *attr, const char *buf, static DEVICE_ATTR_RW(wrap); +static void msc_buffer_unassign(struct msc *msc) +{ + lockdep_assert_held(&msc->buf_mutex); + + if (!msc->mbuf) + return; + + msc->mbuf->unassign(msc->mbuf_priv); + msu_buffer_put(msc->mbuf); + msc->mbuf_priv = NULL; + msc->mbuf = NULL; +} + static ssize_t mode_show(struct device *dev, struct device_attribute *attr, char *buf) { struct msc *msc = dev_get_drvdata(dev); + const char *mode = msc_mode[msc->mode]; + ssize_t ret; + + mutex_lock(&msc->buf_mutex); + if (msc->mbuf) + mode = msc->mbuf->name; + ret = scnprintf(buf, PAGE_SIZE, "%s\n", mode); + mutex_unlock(&msc->buf_mutex); - return scnprintf(buf, PAGE_SIZE, "%s\n", msc_mode[msc->mode]); + return ret; } static ssize_t mode_store(struct device *dev, struct device_attribute *attr, const char *buf, size_t size) { + const struct msu_buffer *mbuf = NULL; struct msc *msc = dev_get_drvdata(dev); size_t len = size; - char *cp; + char *cp, *mode; int i, ret; if (!capable(CAP_SYS_RAWIO)) @@ -1535,17 +1840,59 @@ mode_store(struct device *dev, struct device_attribute *attr, const char *buf, if (cp) len = cp - buf; - for (i = 0; i < ARRAY_SIZE(msc_mode); i++) - if (!strncmp(msc_mode[i], buf, len)) - goto found; + mode = kstrndup(buf, len, GFP_KERNEL); + i = match_string(msc_mode, ARRAY_SIZE(msc_mode), mode); + if (i >= 0) + goto found; + + /* Buffer sinks only work with a usable IRQ */ + if (!msc->do_irq) { + kfree(mode); + return -EINVAL; + } + + mbuf = msu_buffer_get(mode); + kfree(mode); + if (mbuf) + goto found; return -EINVAL; found: mutex_lock(&msc->buf_mutex); + ret = 0; + + /* Same buffer: do nothing */ + if (mbuf && mbuf == msc->mbuf) { + /* put the extra reference we just got */ + msu_buffer_put(mbuf); + goto unlock; + } + ret = msc_buffer_unlocked_free_unless_used(msc); - if (!ret) - msc->mode = i; + if (ret) + goto unlock; + + if (mbuf) { + void *mbuf_priv = mbuf->assign(dev, &i); + + if (!mbuf_priv) { + ret = -ENOMEM; + goto unlock; + } + + msc_buffer_unassign(msc); + msc->mbuf_priv = mbuf_priv; + msc->mbuf = mbuf; + } else { + msc_buffer_unassign(msc); + } + + msc->mode = i; + +unlock: + if (ret && mbuf) + msu_buffer_put(mbuf); mutex_unlock(&msc->buf_mutex); return ret ? ret : size; @@ -1667,7 +2014,12 @@ win_switch_store(struct device *dev, struct device_attribute *attr, return -EINVAL; mutex_lock(&msc->buf_mutex); - if (msc->mode != MSC_MODE_MULTI) + /* + * Window switch can only happen in the "multi" mode. + * If a external buffer is engaged, they have the full + * control over window switching. + */ + if (msc->mode != MSC_MODE_MULTI || msc->mbuf) ret = -ENOTSUPP; else msc_win_switch(msc); @@ -1720,10 +2072,7 @@ static int intel_th_msc_probe(struct intel_th_device *thdev) msc->reg_base = base + msc->index * 0x100; msc->msu_base = base; - err = intel_th_msu_init(msc); - if (err) - return err; - + INIT_WORK(&msc->work, msc_work); err = intel_th_msc_init(msc); if (err) return err; @@ -1739,7 +2088,6 @@ static void intel_th_msc_remove(struct intel_th_device *thdev) int ret; intel_th_msc_deactivate(thdev); - intel_th_msu_deinit(msc); /* * Buffers should not be used at this point except if the diff --git a/drivers/hwtracing/intel_th/msu.h b/drivers/hwtracing/intel_th/msu.h index 574c16004cb2..3f527dd4d727 100644 --- a/drivers/hwtracing/intel_th/msu.h +++ b/drivers/hwtracing/intel_th/msu.h @@ -44,14 +44,6 @@ enum { #define M0BLIE BIT(16) #define M1BLIE BIT(24) -/* MSC operating modes (MSC_MODE) */ -enum { - MSC_MODE_SINGLE = 0, - MSC_MODE_MULTI, - MSC_MODE_EXI, - MSC_MODE_DEBUG, -}; - /* MSCnSTS bits */ #define MSCSTS_WRAPSTAT BIT(1) /* Wrap occurred */ #define MSCSTS_PLE BIT(2) /* Pipeline Empty */ @@ -93,6 +85,16 @@ static inline unsigned long msc_data_sz(struct msc_block_desc *bdesc) return bdesc->valid_dw * 4 - MSC_BDESC; } +static inline unsigned long msc_total_sz(struct msc_block_desc *bdesc) +{ + return bdesc->valid_dw * 4; +} + +static inline unsigned long msc_block_sz(struct msc_block_desc *bdesc) +{ + return bdesc->block_sz * 64 - MSC_BDESC; +} + static inline bool msc_block_wrapped(struct msc_block_desc *bdesc) { if (bdesc->hw_tag & (MSC_HW_TAG_BLOCKWRAP | MSC_HW_TAG_WINWRAP)) @@ -104,7 +106,7 @@ static inline bool msc_block_wrapped(struct msc_block_desc *bdesc) static inline bool msc_block_last_written(struct msc_block_desc *bdesc) { if ((bdesc->hw_tag & MSC_HW_TAG_ENDBIT) || - (msc_data_sz(bdesc) != DATA_IN_PAGE)) + (msc_data_sz(bdesc) != msc_block_sz(bdesc))) return true; return false; diff --git a/include/linux/intel_th.h b/include/linux/intel_th.h new file mode 100644 index 000000000000..9b7f4c22499c --- /dev/null +++ b/include/linux/intel_th.h @@ -0,0 +1,79 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * Intel(R) Trace Hub data structures for implementing buffer sinks. + * + * Copyright (C) 2019 Intel Corporation. + */ + +#ifndef _INTEL_TH_H_ +#define _INTEL_TH_H_ + +#include + +/* MSC operating modes (MSC_MODE) */ +enum { + MSC_MODE_SINGLE = 0, + MSC_MODE_MULTI, + MSC_MODE_EXI, + MSC_MODE_DEBUG, +}; + +struct msu_buffer { + const char *name; + /* + * ->assign() called when buffer 'mode' is set to this driver + * (aka mode_store()) + * @device: struct device * of the msc + * @mode: allows the driver to set HW mode (see the enum above) + * Returns: a pointer to a private structure associated with this + * msc or NULL in case of error. This private structure + * will then be passed into all other callbacks. + */ + void *(*assign)(struct device *dev, int *mode); + /* ->unassign(): some other mode is selected, clean up */ + void (*unassign)(void *priv); + /* + * ->alloc_window(): allocate memory for the window of a given + * size + * @sgt: pointer to sg_table, can be overridden by the buffer + * driver, or kept intact + * Returns: number of sg table entries <= number of pages; + * 0 is treated as an allocation failure. + */ + int (*alloc_window)(void *priv, struct sg_table **sgt, + size_t size); + void (*free_window)(void *priv, struct sg_table *sgt); + /* ->activate(): trace has started */ + void (*activate)(void *priv); + /* ->deactivate(): trace is about to stop */ + void (*deactivate)(void *priv); + /* + * ->ready(): window @sgt is filled up to the last block OR + * tracing is stopped by the user; this window contains + * @bytes data. The window in question transitions into + * the "LOCKED" state, indicating that it can't be used + * by hardware. To clear this state and make the window + * available to the hardware again, call + * intel_th_msc_window_unlock(). + */ + int (*ready)(void *priv, struct sg_table *sgt, size_t bytes); +}; + +int intel_th_msu_buffer_register(const struct msu_buffer *mbuf, + struct module *owner); +void intel_th_msu_buffer_unregister(const struct msu_buffer *mbuf); +void intel_th_msc_window_unlock(struct device *dev, struct sg_table *sgt); + +#define module_intel_th_msu_buffer(__buffer) \ +static int __init __buffer##_init(void) \ +{ \ + return intel_th_msu_buffer_register(&(__buffer), THIS_MODULE); \ +} \ +module_init(__buffer##_init); \ +static void __exit __buffer##_exit(void) \ +{ \ + intel_th_msu_buffer_unregister(&(__buffer)); \ +} \ +module_exit(__buffer##_exit); + +#endif /* _INTEL_TH_H_ */ -- cgit v1.2.3-59-g8ed1b From 509ce4c85bd055ee1013bc853b5d543428b0f017 Mon Sep 17 00:00:00 2001 From: Masahiro Yamada Date: Mon, 29 Jul 2019 00:27:39 +0900 Subject: ppdev: add header include guard Add a header include guard just in case. Signed-off-by: Masahiro Yamada Link: https://lore.kernel.org/r/20190728152739.9249-1-yamada.masahiro@socionext.com Signed-off-by: Greg Kroah-Hartman --- include/uapi/linux/ppdev.h | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/uapi/linux/ppdev.h b/include/uapi/linux/ppdev.h index 8fe3c64d149e..eb895b83f2bd 100644 --- a/include/uapi/linux/ppdev.h +++ b/include/uapi/linux/ppdev.h @@ -15,6 +15,9 @@ * Added PPGETMODES/PPGETMODE/PPGETPHASE, Fred Barnes , 03/01/2001 */ +#ifndef _UAPI_LINUX_PPDEV_H +#define _UAPI_LINUX_PPDEV_H + #define PP_IOCTL 'p' /* Set mode for read/write (e.g. IEEE1284_MODE_EPP) */ @@ -97,4 +100,4 @@ struct ppdev_frob_struct { /* only masks user-visible flags */ #define PP_FLAGMASK (PP_FASTWRITE | PP_FASTREAD | PP_W91284PIC) - +#endif /* _UAPI_LINUX_PPDEV_H */ -- cgit v1.2.3-59-g8ed1b From 69bb18ddfc4331ba1dea9db811caf93e95726408 Mon Sep 17 00:00:00 2001 From: Wu Hao Date: Sun, 4 Aug 2019 18:20:11 +0800 Subject: fpga: dfl: fme: add DFL_FPGA_FME_PORT_RELEASE/ASSIGN ioctl support. In order to support virtualization usage via PCIe SRIOV, this patch adds two ioctls under FPGA Management Engine (FME) to release and assign back the port device. In order to safely turn Port from PF into VF and enable PCIe SRIOV, it requires user to invoke this PORT_RELEASE ioctl to release port firstly to remove userspace interfaces, and then configure the PF/VF access register in FME. After disable SRIOV, it requires user to invoke this PORT_ASSIGN ioctl to attach the port back to PF. Ioctl interfaces: * DFL_FPGA_FME_PORT_RELEASE Release platform device of given port, it deletes port platform device to remove related userspace interfaces on PF. After this function, then it's safe to configure PF/VF access mode to VF, and enable VFs via SRIOV. * DFL_FPGA_FME_PORT_ASSIGN Assign platform device of given port back to PF. After configure PF/VF access mode to PF, this ioctl adds port platform device back to re-enable related userspace interfaces on PF. Signed-off-by: Zhang Yi Z Signed-off-by: Xu Yilun Signed-off-by: Wu Hao Acked-by: Alan Tull Acked-by: Moritz Fischer Signed-off-by: Moritz Fischer Link: https://lore.kernel.org/r/1564914022-3710-2-git-send-email-hao.wu@intel.com Signed-off-by: Greg Kroah-Hartman --- drivers/fpga/dfl-fme-main.c | 42 ++++++++++++++++ drivers/fpga/dfl.c | 113 +++++++++++++++++++++++++++++++++++++----- drivers/fpga/dfl.h | 10 ++++ include/uapi/linux/fpga-dfl.h | 18 +++++++ 4 files changed, 171 insertions(+), 12 deletions(-) (limited to 'include') diff --git a/drivers/fpga/dfl-fme-main.c b/drivers/fpga/dfl-fme-main.c index 0be4635583d5..dfea2dee78c3 100644 --- a/drivers/fpga/dfl-fme-main.c +++ b/drivers/fpga/dfl-fme-main.c @@ -16,6 +16,7 @@ #include #include +#include #include #include "dfl.h" @@ -104,9 +105,50 @@ static void fme_hdr_uinit(struct platform_device *pdev, device_remove_groups(&pdev->dev, fme_hdr_groups); } +static long fme_hdr_ioctl_release_port(struct dfl_feature_platform_data *pdata, + unsigned long arg) +{ + struct dfl_fpga_cdev *cdev = pdata->dfl_cdev; + int port_id; + + if (get_user(port_id, (int __user *)arg)) + return -EFAULT; + + return dfl_fpga_cdev_release_port(cdev, port_id); +} + +static long fme_hdr_ioctl_assign_port(struct dfl_feature_platform_data *pdata, + unsigned long arg) +{ + struct dfl_fpga_cdev *cdev = pdata->dfl_cdev; + int port_id; + + if (get_user(port_id, (int __user *)arg)) + return -EFAULT; + + return dfl_fpga_cdev_assign_port(cdev, port_id); +} + +static long fme_hdr_ioctl(struct platform_device *pdev, + struct dfl_feature *feature, + unsigned int cmd, unsigned long arg) +{ + struct dfl_feature_platform_data *pdata = dev_get_platdata(&pdev->dev); + + switch (cmd) { + case DFL_FPGA_FME_PORT_RELEASE: + return fme_hdr_ioctl_release_port(pdata, arg); + case DFL_FPGA_FME_PORT_ASSIGN: + return fme_hdr_ioctl_assign_port(pdata, arg); + } + + return -ENODEV; +} + static const struct dfl_feature_ops fme_hdr_ops = { .init = fme_hdr_init, .uinit = fme_hdr_uinit, + .ioctl = fme_hdr_ioctl, }; static struct dfl_feature_driver fme_feature_drvs[] = { diff --git a/drivers/fpga/dfl.c b/drivers/fpga/dfl.c index 4b66aaa32b5a..70ffe8b4c157 100644 --- a/drivers/fpga/dfl.c +++ b/drivers/fpga/dfl.c @@ -231,16 +231,20 @@ EXPORT_SYMBOL_GPL(dfl_fpga_port_ops_del); */ int dfl_fpga_check_port_id(struct platform_device *pdev, void *pport_id) { - struct dfl_fpga_port_ops *port_ops = dfl_fpga_port_ops_get(pdev); - int port_id; + struct dfl_feature_platform_data *pdata = dev_get_platdata(&pdev->dev); + struct dfl_fpga_port_ops *port_ops; + + if (pdata->id != FEATURE_DEV_ID_UNUSED) + return pdata->id == *(int *)pport_id; + port_ops = dfl_fpga_port_ops_get(pdev); if (!port_ops || !port_ops->get_id) return 0; - port_id = port_ops->get_id(pdev); + pdata->id = port_ops->get_id(pdev); dfl_fpga_port_ops_put(port_ops); - return port_id == *(int *)pport_id; + return pdata->id == *(int *)pport_id; } EXPORT_SYMBOL_GPL(dfl_fpga_check_port_id); @@ -474,6 +478,7 @@ static int build_info_commit_dev(struct build_feature_devs_info *binfo) pdata->dev = fdev; pdata->num = binfo->feature_num; pdata->dfl_cdev = binfo->cdev; + pdata->id = FEATURE_DEV_ID_UNUSED; mutex_init(&pdata->lock); lockdep_set_class_and_name(&pdata->lock, &dfl_pdata_keys[type], dfl_pdata_key_strings[type]); @@ -973,25 +978,27 @@ void dfl_fpga_feature_devs_remove(struct dfl_fpga_cdev *cdev) { struct dfl_feature_platform_data *pdata, *ptmp; - remove_feature_devs(cdev); - mutex_lock(&cdev->lock); - if (cdev->fme_dev) { - /* the fme should be unregistered. */ - WARN_ON(device_is_registered(cdev->fme_dev)); + if (cdev->fme_dev) put_device(cdev->fme_dev); - } list_for_each_entry_safe(pdata, ptmp, &cdev->port_dev_list, node) { struct platform_device *port_dev = pdata->dev; - /* the port should be unregistered. */ - WARN_ON(device_is_registered(&port_dev->dev)); + /* remove released ports */ + if (!device_is_registered(&port_dev->dev)) { + dfl_id_free(feature_dev_id_type(port_dev), + port_dev->id); + platform_device_put(port_dev); + } + list_del(&pdata->node); put_device(&port_dev->dev); } mutex_unlock(&cdev->lock); + remove_feature_devs(cdev); + fpga_region_unregister(cdev->region); devm_kfree(cdev->parent, cdev); } @@ -1042,6 +1049,88 @@ static int __init dfl_fpga_init(void) return ret; } +/** + * dfl_fpga_cdev_release_port - release a port platform device + * + * @cdev: parent container device. + * @port_id: id of the port platform device. + * + * This function allows user to release a port platform device. This is a + * mandatory step before turn a port from PF into VF for SRIOV support. + * + * Return: 0 on success, negative error code otherwise. + */ +int dfl_fpga_cdev_release_port(struct dfl_fpga_cdev *cdev, int port_id) +{ + struct platform_device *port_pdev; + int ret = -ENODEV; + + mutex_lock(&cdev->lock); + port_pdev = __dfl_fpga_cdev_find_port(cdev, &port_id, + dfl_fpga_check_port_id); + if (!port_pdev) + goto unlock_exit; + + if (!device_is_registered(&port_pdev->dev)) { + ret = -EBUSY; + goto put_dev_exit; + } + + ret = dfl_feature_dev_use_begin(dev_get_platdata(&port_pdev->dev)); + if (ret) + goto put_dev_exit; + + platform_device_del(port_pdev); + cdev->released_port_num++; +put_dev_exit: + put_device(&port_pdev->dev); +unlock_exit: + mutex_unlock(&cdev->lock); + return ret; +} +EXPORT_SYMBOL_GPL(dfl_fpga_cdev_release_port); + +/** + * dfl_fpga_cdev_assign_port - assign a port platform device back + * + * @cdev: parent container device. + * @port_id: id of the port platform device. + * + * This function allows user to assign a port platform device back. This is + * a mandatory step after disable SRIOV support. + * + * Return: 0 on success, negative error code otherwise. + */ +int dfl_fpga_cdev_assign_port(struct dfl_fpga_cdev *cdev, int port_id) +{ + struct platform_device *port_pdev; + int ret = -ENODEV; + + mutex_lock(&cdev->lock); + port_pdev = __dfl_fpga_cdev_find_port(cdev, &port_id, + dfl_fpga_check_port_id); + if (!port_pdev) + goto unlock_exit; + + if (device_is_registered(&port_pdev->dev)) { + ret = -EBUSY; + goto put_dev_exit; + } + + ret = platform_device_add(port_pdev); + if (ret) + goto put_dev_exit; + + dfl_feature_dev_use_end(dev_get_platdata(&port_pdev->dev)); + cdev->released_port_num--; +put_dev_exit: + put_device(&port_pdev->dev); +unlock_exit: + mutex_unlock(&cdev->lock); + return ret; +} +EXPORT_SYMBOL_GPL(dfl_fpga_cdev_assign_port); + static void __exit dfl_fpga_exit(void) { dfl_chardev_uinit(); diff --git a/drivers/fpga/dfl.h b/drivers/fpga/dfl.h index a8b869e9e5b7..6f7855e57869 100644 --- a/drivers/fpga/dfl.h +++ b/drivers/fpga/dfl.h @@ -183,6 +183,8 @@ struct dfl_feature { #define DEV_STATUS_IN_USE 0 +#define FEATURE_DEV_ID_UNUSED (-1) + /** * struct dfl_feature_platform_data - platform data for feature devices * @@ -191,6 +193,7 @@ struct dfl_feature { * @cdev: cdev of feature dev. * @dev: ptr to platform device linked with this platform data. * @dfl_cdev: ptr to container device. + * @id: id used for this feature device. * @disable_count: count for port disable. * @num: number for sub features. * @dev_status: dev status (e.g. DEV_STATUS_IN_USE). @@ -203,6 +206,7 @@ struct dfl_feature_platform_data { struct cdev cdev; struct platform_device *dev; struct dfl_fpga_cdev *dfl_cdev; + int id; unsigned int disable_count; unsigned long dev_status; void *private; @@ -373,6 +377,7 @@ void dfl_fpga_enum_info_free(struct dfl_fpga_enum_info *info); * @fme_dev: FME feature device under this container device. * @lock: mutex lock to protect the port device list. * @port_dev_list: list of all port feature devices under this container device. + * @released_port_num: released port number under this container device. */ struct dfl_fpga_cdev { struct device *parent; @@ -380,6 +385,7 @@ struct dfl_fpga_cdev { struct device *fme_dev; struct mutex lock; struct list_head port_dev_list; + int released_port_num; }; struct dfl_fpga_cdev * @@ -407,4 +413,8 @@ dfl_fpga_cdev_find_port(struct dfl_fpga_cdev *cdev, void *data, return pdev; } + +int dfl_fpga_cdev_release_port(struct dfl_fpga_cdev *cdev, int port_id); +int dfl_fpga_cdev_assign_port(struct dfl_fpga_cdev *cdev, int port_id); + #endif /* __FPGA_DFL_H */ diff --git a/include/uapi/linux/fpga-dfl.h b/include/uapi/linux/fpga-dfl.h index 2e324e515c41..ec70a0746e59 100644 --- a/include/uapi/linux/fpga-dfl.h +++ b/include/uapi/linux/fpga-dfl.h @@ -176,4 +176,22 @@ struct dfl_fpga_fme_port_pr { #define DFL_FPGA_FME_PORT_PR _IO(DFL_FPGA_MAGIC, DFL_FME_BASE + 0) +/** + * DFL_FPGA_FME_PORT_RELEASE - _IOW(DFL_FPGA_MAGIC, DFL_FME_BASE + 1, + * int port_id) + * + * Driver releases the port per Port ID provided by caller. + * Return: 0 on success, -errno on failure. + */ +#define DFL_FPGA_FME_PORT_RELEASE _IOW(DFL_FPGA_MAGIC, DFL_FME_BASE + 1, int) + +/** + * DFL_FPGA_FME_PORT_ASSIGN - _IOW(DFL_FPGA_MAGIC, DFL_FME_BASE + 2, + * int port_id) + * + * Driver assigns the port back per Port ID provided by caller. + * Return: 0 on success, -errno on failure. + */ +#define DFL_FPGA_FME_PORT_ASSIGN _IOW(DFL_FPGA_MAGIC, DFL_FME_BASE + 2, int) + #endif /* _UAPI_LINUX_FPGA_DFL_H */ -- cgit v1.2.3-59-g8ed1b From 127ab2cc5f19692efe422935267b9db0845b2b04 Mon Sep 17 00:00:00 2001 From: Georgi Djakov Date: Fri, 9 Aug 2019 15:13:23 +0300 Subject: interconnect: Add support for path tags Consumers may have use cases with different bandwidth requirements based on the system or driver state. The consumer driver can append a specific tag to the path and pass this information to the interconnect platform driver to do the aggregation based on this state. Introduce icc_set_tag() function that will allow the consumers to append an optional tag to each path. The aggregation of these tagged paths is platform specific. Reviewed-by: Evan Green Signed-off-by: Georgi Djakov --- drivers/interconnect/core.c | 24 +++++++++++++++++++++++- drivers/interconnect/qcom/sdm845.c | 2 +- include/linux/interconnect-provider.h | 4 ++-- include/linux/interconnect.h | 5 +++++ 4 files changed, 31 insertions(+), 4 deletions(-) (limited to 'include') diff --git a/drivers/interconnect/core.c b/drivers/interconnect/core.c index 871eb4bc4efc..251354bb7fdc 100644 --- a/drivers/interconnect/core.c +++ b/drivers/interconnect/core.c @@ -29,6 +29,7 @@ static struct dentry *icc_debugfs_dir; * @req_node: entry in list of requests for the particular @node * @node: the interconnect node to which this constraint applies * @dev: reference to the device that sets the constraints + * @tag: path tag (optional) * @avg_bw: an integer describing the average bandwidth in kBps * @peak_bw: an integer describing the peak bandwidth in kBps */ @@ -36,6 +37,7 @@ struct icc_req { struct hlist_node req_node; struct icc_node *node; struct device *dev; + u32 tag; u32 avg_bw; u32 peak_bw; }; @@ -204,7 +206,7 @@ static int aggregate_requests(struct icc_node *node) node->peak_bw = 0; hlist_for_each_entry(r, &node->req_list, req_node) - p->aggregate(node, r->avg_bw, r->peak_bw, + p->aggregate(node, r->tag, r->avg_bw, r->peak_bw, &node->avg_bw, &node->peak_bw); return 0; @@ -385,6 +387,26 @@ struct icc_path *of_icc_get(struct device *dev, const char *name) } EXPORT_SYMBOL_GPL(of_icc_get); +/** + * icc_set_tag() - set an optional tag on a path + * @path: the path we want to tag + * @tag: the tag value + * + * This function allows consumers to append a tag to the requests associated + * with a path, so that a different aggregation could be done based on this tag. + */ +void icc_set_tag(struct icc_path *path, u32 tag) +{ + int i; + + if (!path) + return; + + for (i = 0; i < path->num_nodes; i++) + path->reqs[i].tag = tag; +} +EXPORT_SYMBOL_GPL(icc_set_tag); + /** * icc_set_bw() - set bandwidth constraints on an interconnect path * @path: reference to the path returned by icc_get() diff --git a/drivers/interconnect/qcom/sdm845.c b/drivers/interconnect/qcom/sdm845.c index 4915b78da673..fb526004c82e 100644 --- a/drivers/interconnect/qcom/sdm845.c +++ b/drivers/interconnect/qcom/sdm845.c @@ -626,7 +626,7 @@ static void bcm_aggregate(struct qcom_icc_bcm *bcm) bcm->dirty = false; } -static int qcom_icc_aggregate(struct icc_node *node, u32 avg_bw, +static int qcom_icc_aggregate(struct icc_node *node, u32 tag, u32 avg_bw, u32 peak_bw, u32 *agg_avg, u32 *agg_peak) { size_t i; diff --git a/include/linux/interconnect-provider.h b/include/linux/interconnect-provider.h index 63caccadc2db..4ee19fd41568 100644 --- a/include/linux/interconnect-provider.h +++ b/include/linux/interconnect-provider.h @@ -45,8 +45,8 @@ struct icc_provider { struct list_head provider_list; struct list_head nodes; int (*set)(struct icc_node *src, struct icc_node *dst); - int (*aggregate)(struct icc_node *node, u32 avg_bw, u32 peak_bw, - u32 *agg_avg, u32 *agg_peak); + int (*aggregate)(struct icc_node *node, u32 tag, u32 avg_bw, + u32 peak_bw, u32 *agg_avg, u32 *agg_peak); struct icc_node* (*xlate)(struct of_phandle_args *spec, void *data); struct device *dev; int users; diff --git a/include/linux/interconnect.h b/include/linux/interconnect.h index dc25864755ba..d70a914cba11 100644 --- a/include/linux/interconnect.h +++ b/include/linux/interconnect.h @@ -30,6 +30,7 @@ struct icc_path *icc_get(struct device *dev, const int src_id, struct icc_path *of_icc_get(struct device *dev, const char *name); void icc_put(struct icc_path *path); int icc_set_bw(struct icc_path *path, u32 avg_bw, u32 peak_bw); +void icc_set_tag(struct icc_path *path, u32 tag); #else @@ -54,6 +55,10 @@ static inline int icc_set_bw(struct icc_path *path, u32 avg_bw, u32 peak_bw) return 0; } +static inline void icc_set_tag(struct icc_path *path, u32 tag) +{ +} + #endif /* CONFIG_INTERCONNECT */ #endif /* __LINUX_INTERCONNECT_H */ -- cgit v1.2.3-59-g8ed1b From cbd5a9c28bb5d2560be34fc6a2b6e60197628433 Mon Sep 17 00:00:00 2001 From: Georgi Djakov Date: Fri, 9 Aug 2019 15:13:24 +0300 Subject: interconnect: Add pre_aggregate() callback Introduce an optional callback in interconnect provider drivers. It can be used for implementing actions, that need to be executed before the actual aggregation of the bandwidth requests has started. The benefit of this for now is that it will significantly simplify the code in provider drivers. Suggested-by: Evan Green Reviewed-by: Evan Green Signed-off-by: Georgi Djakov --- drivers/interconnect/core.c | 3 +++ include/linux/interconnect-provider.h | 3 +++ 2 files changed, 6 insertions(+) (limited to 'include') diff --git a/drivers/interconnect/core.c b/drivers/interconnect/core.c index 251354bb7fdc..7b971228df38 100644 --- a/drivers/interconnect/core.c +++ b/drivers/interconnect/core.c @@ -205,6 +205,9 @@ static int aggregate_requests(struct icc_node *node) node->avg_bw = 0; node->peak_bw = 0; + if (p->pre_aggregate) + p->pre_aggregate(node); + hlist_for_each_entry(r, &node->req_list, req_node) p->aggregate(node, r->tag, r->avg_bw, r->peak_bw, &node->avg_bw, &node->peak_bw); diff --git a/include/linux/interconnect-provider.h b/include/linux/interconnect-provider.h index 4ee19fd41568..b16f9effa555 100644 --- a/include/linux/interconnect-provider.h +++ b/include/linux/interconnect-provider.h @@ -36,6 +36,8 @@ struct icc_node *of_icc_xlate_onecell(struct of_phandle_args *spec, * @nodes: internal list of the interconnect provider nodes * @set: pointer to device specific set operation function * @aggregate: pointer to device specific aggregate operation function + * @pre_aggregate: pointer to device specific function that is called + * before the aggregation begins (optional) * @xlate: provider-specific callback for mapping nodes from phandle arguments * @dev: the device this interconnect provider belongs to * @users: count of active users @@ -47,6 +49,7 @@ struct icc_provider { int (*set)(struct icc_node *src, struct icc_node *dst); int (*aggregate)(struct icc_node *node, u32 tag, u32 avg_bw, u32 peak_bw, u32 *agg_avg, u32 *agg_peak); + void (*pre_aggregate)(struct icc_node *node); struct icc_node* (*xlate)(struct of_phandle_args *spec, void *data); struct device *dev; int users; -- cgit v1.2.3-59-g8ed1b From 24f516ebbab8a212a9aa8c3d69f185371f5e200b Mon Sep 17 00:00:00 2001 From: Georgi Djakov Date: Tue, 23 Jul 2019 17:23:35 +0300 Subject: dt-bindings: interconnect: Add Qualcomm QCS404 DT bindings The Qualcomm QCS404 platform has several buses that could be controlled and tuned according to the bandwidth demand. Reviewed-by: Bjorn Andersson Reviewed-by: Rob Herring Signed-off-by: Georgi Djakov --- .../bindings/interconnect/qcom,qcs404.txt | 45 +++++++++++ include/dt-bindings/interconnect/qcom,qcs404.h | 88 ++++++++++++++++++++++ 2 files changed, 133 insertions(+) create mode 100644 Documentation/devicetree/bindings/interconnect/qcom,qcs404.txt create mode 100644 include/dt-bindings/interconnect/qcom,qcs404.h (limited to 'include') diff --git a/Documentation/devicetree/bindings/interconnect/qcom,qcs404.txt b/Documentation/devicetree/bindings/interconnect/qcom,qcs404.txt new file mode 100644 index 000000000000..c07d89812b73 --- /dev/null +++ b/Documentation/devicetree/bindings/interconnect/qcom,qcs404.txt @@ -0,0 +1,45 @@ +Qualcomm QCS404 Network-On-Chip interconnect driver binding +----------------------------------------------------------- + +Required properties : +- compatible : shall contain only one of the following: + "qcom,qcs404-bimc" + "qcom,qcs404-pcnoc" + "qcom,qcs404-snoc" +- #interconnect-cells : should contain 1 + +reg : specifies the physical base address and size of registers +clocks : list of phandles and specifiers to all interconnect bus clocks +clock-names : clock names should include both "bus" and "bus_a" + +Example: + +soc { + ... + bimc: interconnect@400000 { + reg = <0x00400000 0x80000>; + compatible = "qcom,qcs404-bimc"; + #interconnect-cells = <1>; + clock-names = "bus", "bus_a"; + clocks = <&rpmcc RPM_SMD_BIMC_CLK>, + <&rpmcc RPM_SMD_BIMC_A_CLK>; + }; + + pnoc: interconnect@500000 { + reg = <0x00500000 0x15080>; + compatible = "qcom,qcs404-pcnoc"; + #interconnect-cells = <1>; + clock-names = "bus", "bus_a"; + clocks = <&rpmcc RPM_SMD_PNOC_CLK>, + <&rpmcc RPM_SMD_PNOC_A_CLK>; + }; + + snoc: interconnect@580000 { + reg = <0x00580000 0x23080>; + compatible = "qcom,qcs404-snoc"; + #interconnect-cells = <1>; + clock-names = "bus", "bus_a"; + clocks = <&rpmcc RPM_SMD_SNOC_CLK>, + <&rpmcc RPM_SMD_SNOC_A_CLK>; + }; +}; diff --git a/include/dt-bindings/interconnect/qcom,qcs404.h b/include/dt-bindings/interconnect/qcom,qcs404.h new file mode 100644 index 000000000000..960f6e39c5f2 --- /dev/null +++ b/include/dt-bindings/interconnect/qcom,qcs404.h @@ -0,0 +1,88 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * Qualcomm interconnect IDs + * + * Copyright (c) 2019, Linaro Ltd. + * Author: Georgi Djakov + */ + +#ifndef __DT_BINDINGS_INTERCONNECT_QCOM_QCS404_H +#define __DT_BINDINGS_INTERCONNECT_QCOM_QCS404_H + +#define MASTER_AMPSS_M0 0 +#define MASTER_OXILI 1 +#define MASTER_MDP_PORT0 2 +#define MASTER_SNOC_BIMC_1 3 +#define MASTER_TCU_0 4 +#define SLAVE_EBI_CH0 5 +#define SLAVE_BIMC_SNOC 6 + +#define MASTER_SPDM 0 +#define MASTER_BLSP_1 1 +#define MASTER_BLSP_2 2 +#define MASTER_XI_USB_HS1 3 +#define MASTER_CRYPT0 4 +#define MASTER_SDCC_1 5 +#define MASTER_SDCC_2 6 +#define MASTER_SNOC_PCNOC 7 +#define MASTER_QPIC 8 +#define PCNOC_INT_0 9 +#define PCNOC_INT_2 10 +#define PCNOC_INT_3 11 +#define PCNOC_S_0 12 +#define PCNOC_S_1 13 +#define PCNOC_S_2 14 +#define PCNOC_S_3 15 +#define PCNOC_S_4 16 +#define PCNOC_S_6 17 +#define PCNOC_S_7 18 +#define PCNOC_S_8 19 +#define PCNOC_S_9 20 +#define PCNOC_S_10 21 +#define PCNOC_S_11 22 +#define SLAVE_SPDM 23 +#define SLAVE_PDM 24 +#define SLAVE_PRNG 25 +#define SLAVE_TCSR 26 +#define SLAVE_SNOC_CFG 27 +#define SLAVE_MESSAGE_RAM 28 +#define SLAVE_DISP_SS_CFG 29 +#define SLAVE_GPU_CFG 30 +#define SLAVE_BLSP_1 31 +#define SLAVE_BLSP_2 32 +#define SLAVE_TLMM_NORTH 33 +#define SLAVE_PCIE 34 +#define SLAVE_ETHERNET 35 +#define SLAVE_TLMM_EAST 36 +#define SLAVE_TCU 37 +#define SLAVE_PMIC_ARB 38 +#define SLAVE_SDCC_1 39 +#define SLAVE_SDCC_2 40 +#define SLAVE_TLMM_SOUTH 41 +#define SLAVE_USB_HS 42 +#define SLAVE_USB3 43 +#define SLAVE_CRYPTO_0_CFG 44 +#define SLAVE_PCNOC_SNOC 45 + +#define MASTER_QDSS_BAM 0 +#define MASTER_BIMC_SNOC 1 +#define MASTER_PCNOC_SNOC 2 +#define MASTER_QDSS_ETR 3 +#define MASTER_EMAC 4 +#define MASTER_PCIE 5 +#define MASTER_USB3 6 +#define QDSS_INT 7 +#define SNOC_INT_0 8 +#define SNOC_INT_1 9 +#define SNOC_INT_2 10 +#define SLAVE_KPSS_AHB 11 +#define SLAVE_WCSS 12 +#define SLAVE_SNOC_BIMC_1 13 +#define SLAVE_IMEM 14 +#define SLAVE_SNOC_PCNOC 15 +#define SLAVE_QDSS_STM 16 +#define SLAVE_CATS_0 17 +#define SLAVE_CATS_1 18 +#define SLAVE_LPASS 19 + +#endif -- cgit v1.2.3-59-g8ed1b From 6d54e455689edcf1f9ef30761dd4fdfdc1cba33a Mon Sep 17 00:00:00 2001 From: Dragan Cvetic Date: Sat, 27 Jul 2019 09:33:51 +0100 Subject: misc: xilinx_sdfec: Store driver config and state Stores configuration based on parameters from the DT node and values from the SD-FEC core plus reads the default state from the SD-FEC core. To obtain values from the core register read, write capabilities have been added plus related register map details. Tested-by: Dragan Cvetic Signed-off-by: Derek Kiernan Signed-off-by: Dragan Cvetic Link: https://lore.kernel.org/r/1564216438-322406-2-git-send-email-dragan.cvetic@xilinx.com Signed-off-by: Greg Kroah-Hartman --- drivers/misc/xilinx_sdfec.c | 305 ++++++++++++++++++++++++++++++++++++++- include/uapi/misc/xilinx_sdfec.h | 138 ++++++++++++++++++ 2 files changed, 437 insertions(+), 6 deletions(-) create mode 100644 include/uapi/misc/xilinx_sdfec.h (limited to 'include') diff --git a/drivers/misc/xilinx_sdfec.c b/drivers/misc/xilinx_sdfec.c index f257d3812110..24d9f79fe073 100644 --- a/drivers/misc/xilinx_sdfec.c +++ b/drivers/misc/xilinx_sdfec.c @@ -20,11 +20,92 @@ #include #include +#include + #define DEV_NAME_LEN 12 static struct idr dev_idr; static struct mutex dev_idr_lock; +/* Xilinx SDFEC Register Map */ +/* CODE_WRI_PROTECT Register */ +#define XSDFEC_CODE_WR_PROTECT_ADDR (0x4) + +/* ACTIVE Register */ +#define XSDFEC_ACTIVE_ADDR (0x8) +#define XSDFEC_IS_ACTIVITY_SET (0x1) + +/* AXIS_WIDTH Register */ +#define XSDFEC_AXIS_WIDTH_ADDR (0xC) +#define XSDFEC_AXIS_DOUT_WORDS_LSB (5) +#define XSDFEC_AXIS_DOUT_WIDTH_LSB (3) +#define XSDFEC_AXIS_DIN_WORDS_LSB (2) +#define XSDFEC_AXIS_DIN_WIDTH_LSB (0) + +/* AXIS_ENABLE Register */ +#define XSDFEC_AXIS_ENABLE_ADDR (0x10) +#define XSDFEC_AXIS_OUT_ENABLE_MASK (0x38) +#define XSDFEC_AXIS_IN_ENABLE_MASK (0x7) +#define XSDFEC_AXIS_ENABLE_MASK \ + (XSDFEC_AXIS_OUT_ENABLE_MASK | XSDFEC_AXIS_IN_ENABLE_MASK) + +/* FEC_CODE Register */ +#define XSDFEC_FEC_CODE_ADDR (0x14) + +/* ORDER Register Map */ +#define XSDFEC_ORDER_ADDR (0x18) + +/* Interrupt Status Register */ +#define XSDFEC_ISR_ADDR (0x1C) +/* Interrupt Status Register Bit Mask */ +#define XSDFEC_ISR_MASK (0x3F) + +/* Write Only - Interrupt Enable Register */ +#define XSDFEC_IER_ADDR (0x20) +/* Write Only - Interrupt Disable Register */ +#define XSDFEC_IDR_ADDR (0x24) +/* Read Only - Interrupt Mask Register */ +#define XSDFEC_IMR_ADDR (0x28) + +/* ECC Interrupt Status Register */ +#define XSDFEC_ECC_ISR_ADDR (0x2C) +/* Single Bit Errors */ +#define XSDFEC_ECC_ISR_SBE_MASK (0x7FF) +/* PL Initialize Single Bit Errors */ +#define XSDFEC_PL_INIT_ECC_ISR_SBE_MASK (0x3C00000) +/* Multi Bit Errors */ +#define XSDFEC_ECC_ISR_MBE_MASK (0x3FF800) +/* PL Initialize Multi Bit Errors */ +#define XSDFEC_PL_INIT_ECC_ISR_MBE_MASK (0x3C000000) +/* Multi Bit Error to Event Shift */ +#define XSDFEC_ECC_ISR_MBE_TO_EVENT_SHIFT (11) +/* PL Initialize Multi Bit Error to Event Shift */ +#define XSDFEC_PL_INIT_ECC_ISR_MBE_TO_EVENT_SHIFT (4) +/* ECC Interrupt Status Bit Mask */ +#define XSDFEC_ECC_ISR_MASK (XSDFEC_ECC_ISR_SBE_MASK | XSDFEC_ECC_ISR_MBE_MASK) +/* ECC Interrupt Status PL Initialize Bit Mask */ +#define XSDFEC_PL_INIT_ECC_ISR_MASK \ + (XSDFEC_PL_INIT_ECC_ISR_SBE_MASK | XSDFEC_PL_INIT_ECC_ISR_MBE_MASK) +/* ECC Interrupt Status All Bit Mask */ +#define XSDFEC_ALL_ECC_ISR_MASK \ + (XSDFEC_ECC_ISR_MASK | XSDFEC_PL_INIT_ECC_ISR_MASK) +/* ECC Interrupt Status Single Bit Errors Mask */ +#define XSDFEC_ALL_ECC_ISR_SBE_MASK \ + (XSDFEC_ECC_ISR_SBE_MASK | XSDFEC_PL_INIT_ECC_ISR_SBE_MASK) +/* ECC Interrupt Status Multi Bit Errors Mask */ +#define XSDFEC_ALL_ECC_ISR_MBE_MASK \ + (XSDFEC_ECC_ISR_MBE_MASK | XSDFEC_PL_INIT_ECC_ISR_MBE_MASK) + +/* Write Only - ECC Interrupt Enable Register */ +#define XSDFEC_ECC_IER_ADDR (0x30) +/* Write Only - ECC Interrupt Disable Register */ +#define XSDFEC_ECC_IDR_ADDR (0x34) +/* Read Only - ECC Interrupt Mask Register */ +#define XSDFEC_ECC_IMR_ADDR (0x38) + +/* BYPASS Register */ +#define XSDFEC_BYPASS_ADDR (0x3C) + /** * struct xsdfec_clks - For managing SD-FEC clocks * @core_clk: Main processing clock for core @@ -49,31 +130,237 @@ struct xsdfec_clks { /** * struct xsdfec_dev - Driver data for SDFEC - * @regs: device physical base address - * @dev: pointer to device struct * @miscdev: Misc device handle - * @error_data_lock: Error counter and states spinlock * @clks: Clocks managed by the SDFEC driver + * @regs: device physical base address + * @dev: pointer to device struct + * @config: Configuration of the SDFEC device * @dev_name: Device name + * @state: State of the SDFEC device + * @error_data_lock: Error counter and states spinlock * @dev_id: Device ID * * This structure contains necessary state for SDFEC driver to operate */ struct xsdfec_dev { + struct miscdevice miscdev; + struct xsdfec_clks clks; void __iomem *regs; struct device *dev; - struct miscdevice miscdev; + struct xsdfec_config config; + char dev_name[DEV_NAME_LEN]; + enum xsdfec_state state; /* Spinlock to protect state_updated and stats_updated */ spinlock_t error_data_lock; - struct xsdfec_clks clks; - char dev_name[DEV_NAME_LEN]; int dev_id; }; +static inline void xsdfec_regwrite(struct xsdfec_dev *xsdfec, u32 addr, + u32 value) +{ + dev_dbg(xsdfec->dev, "Writing 0x%x to offset 0x%x", value, addr); + iowrite32(value, xsdfec->regs + addr); +} + +static inline u32 xsdfec_regread(struct xsdfec_dev *xsdfec, u32 addr) +{ + u32 rval; + + rval = ioread32(xsdfec->regs + addr); + dev_dbg(xsdfec->dev, "Read value = 0x%x from offset 0x%x", rval, addr); + return rval; +} + +static void update_bool_config_from_reg(struct xsdfec_dev *xsdfec, + u32 reg_offset, u32 bit_num, + char *config_value) +{ + u32 reg_val; + u32 bit_mask = 1 << bit_num; + + reg_val = xsdfec_regread(xsdfec, reg_offset); + *config_value = (reg_val & bit_mask) > 0; +} + +static void update_config_from_hw(struct xsdfec_dev *xsdfec) +{ + u32 reg_value; + bool sdfec_started; + + /* Update the Order */ + reg_value = xsdfec_regread(xsdfec, XSDFEC_ORDER_ADDR); + xsdfec->config.order = reg_value; + + update_bool_config_from_reg(xsdfec, XSDFEC_BYPASS_ADDR, + 0, /* Bit Number, maybe change to mask */ + &xsdfec->config.bypass); + + update_bool_config_from_reg(xsdfec, XSDFEC_CODE_WR_PROTECT_ADDR, + 0, /* Bit Number */ + &xsdfec->config.code_wr_protect); + + reg_value = xsdfec_regread(xsdfec, XSDFEC_IMR_ADDR); + xsdfec->config.irq.enable_isr = (reg_value & XSDFEC_ISR_MASK) > 0; + + reg_value = xsdfec_regread(xsdfec, XSDFEC_ECC_IMR_ADDR); + xsdfec->config.irq.enable_ecc_isr = + (reg_value & XSDFEC_ECC_ISR_MASK) > 0; + + reg_value = xsdfec_regread(xsdfec, XSDFEC_AXIS_ENABLE_ADDR); + sdfec_started = (reg_value & XSDFEC_AXIS_IN_ENABLE_MASK) > 0; + if (sdfec_started) + xsdfec->state = XSDFEC_STARTED; + else + xsdfec->state = XSDFEC_STOPPED; +} + +static u32 +xsdfec_translate_axis_width_cfg_val(enum xsdfec_axis_width axis_width_cfg) +{ + u32 axis_width_field = 0; + + switch (axis_width_cfg) { + case XSDFEC_1x128b: + axis_width_field = 0; + break; + case XSDFEC_2x128b: + axis_width_field = 1; + break; + case XSDFEC_4x128b: + axis_width_field = 2; + break; + } + + return axis_width_field; +} + +static u32 xsdfec_translate_axis_words_cfg_val(enum xsdfec_axis_word_include + axis_word_inc_cfg) +{ + u32 axis_words_field = 0; + + if (axis_word_inc_cfg == XSDFEC_FIXED_VALUE || + axis_word_inc_cfg == XSDFEC_IN_BLOCK) + axis_words_field = 0; + else if (axis_word_inc_cfg == XSDFEC_PER_AXI_TRANSACTION) + axis_words_field = 1; + + return axis_words_field; +} + +static int xsdfec_cfg_axi_streams(struct xsdfec_dev *xsdfec) +{ + u32 reg_value; + u32 dout_words_field; + u32 dout_width_field; + u32 din_words_field; + u32 din_width_field; + struct xsdfec_config *config = &xsdfec->config; + + /* translate config info to register values */ + dout_words_field = + xsdfec_translate_axis_words_cfg_val(config->dout_word_include); + dout_width_field = + xsdfec_translate_axis_width_cfg_val(config->dout_width); + din_words_field = + xsdfec_translate_axis_words_cfg_val(config->din_word_include); + din_width_field = + xsdfec_translate_axis_width_cfg_val(config->din_width); + + reg_value = dout_words_field << XSDFEC_AXIS_DOUT_WORDS_LSB; + reg_value |= dout_width_field << XSDFEC_AXIS_DOUT_WIDTH_LSB; + reg_value |= din_words_field << XSDFEC_AXIS_DIN_WORDS_LSB; + reg_value |= din_width_field << XSDFEC_AXIS_DIN_WIDTH_LSB; + + xsdfec_regwrite(xsdfec, XSDFEC_AXIS_WIDTH_ADDR, reg_value); + + return 0; +} + static const struct file_operations xsdfec_fops = { .owner = THIS_MODULE, }; +static int xsdfec_parse_of(struct xsdfec_dev *xsdfec) +{ + struct device *dev = xsdfec->dev; + struct device_node *node = dev->of_node; + int rval; + const char *fec_code; + u32 din_width; + u32 din_word_include; + u32 dout_width; + u32 dout_word_include; + + rval = of_property_read_string(node, "xlnx,sdfec-code", &fec_code); + if (rval < 0) + return rval; + + if (!strcasecmp(fec_code, "ldpc")) + xsdfec->config.code = XSDFEC_LDPC_CODE; + else if (!strcasecmp(fec_code, "turbo")) + xsdfec->config.code = XSDFEC_TURBO_CODE; + else + return -EINVAL; + + rval = of_property_read_u32(node, "xlnx,sdfec-din-words", + &din_word_include); + if (rval < 0) + return rval; + + if (din_word_include < XSDFEC_AXIS_WORDS_INCLUDE_MAX) + xsdfec->config.din_word_include = din_word_include; + else + return -EINVAL; + + rval = of_property_read_u32(node, "xlnx,sdfec-din-width", &din_width); + if (rval < 0) + return rval; + + switch (din_width) { + /* Fall through and set for valid values */ + case XSDFEC_1x128b: + case XSDFEC_2x128b: + case XSDFEC_4x128b: + xsdfec->config.din_width = din_width; + break; + default: + return -EINVAL; + } + + rval = of_property_read_u32(node, "xlnx,sdfec-dout-words", + &dout_word_include); + if (rval < 0) + return rval; + + if (dout_word_include < XSDFEC_AXIS_WORDS_INCLUDE_MAX) + xsdfec->config.dout_word_include = dout_word_include; + else + return -EINVAL; + + rval = of_property_read_u32(node, "xlnx,sdfec-dout-width", &dout_width); + if (rval < 0) + return rval; + + switch (dout_width) { + /* Fall through and set for valid values */ + case XSDFEC_1x128b: + case XSDFEC_2x128b: + case XSDFEC_4x128b: + xsdfec->config.dout_width = dout_width; + break; + default: + return -EINVAL; + } + + /* Write LDPC to CODE Register */ + xsdfec_regwrite(xsdfec, XSDFEC_FEC_CODE_ADDR, xsdfec->config.code); + + xsdfec_cfg_axi_streams(xsdfec); + + return 0; +} + static int xsdfec_clk_init(struct platform_device *pdev, struct xsdfec_clks *clks) { @@ -260,6 +547,12 @@ static int xsdfec_probe(struct platform_device *pdev) goto err_xsdfec_dev; } + err = xsdfec_parse_of(xsdfec); + if (err < 0) + goto err_xsdfec_dev; + + update_config_from_hw(xsdfec); + /* Save driver private data */ platform_set_drvdata(pdev, xsdfec); diff --git a/include/uapi/misc/xilinx_sdfec.h b/include/uapi/misc/xilinx_sdfec.h new file mode 100644 index 000000000000..330ea25ddfa1 --- /dev/null +++ b/include/uapi/misc/xilinx_sdfec.h @@ -0,0 +1,138 @@ +/* SPDX-License-Identifier: GPL-2.0+ WITH Linux-syscall-note */ +/* + * Xilinx SD-FEC + * + * Copyright (C) 2019 Xilinx, Inc. + * + * Description: + * This driver is developed for SDFEC16 IP. It provides a char device + * in sysfs and supports file operations like open(), close() and ioctl(). + */ +#ifndef __XILINX_SDFEC_H__ +#define __XILINX_SDFEC_H__ + +#include + +/** + * enum xsdfec_code - Code Type. + * @XSDFEC_TURBO_CODE: Driver is configured for Turbo mode. + * @XSDFEC_LDPC_CODE: Driver is configured for LDPC mode. + * + * This enum is used to indicate the mode of the driver. The mode is determined + * by checking which codes are set in the driver. Note that the mode cannot be + * changed by the driver. + */ +enum xsdfec_code { + XSDFEC_TURBO_CODE = 0, + XSDFEC_LDPC_CODE, +}; + +/** + * enum xsdfec_order - Order + * @XSDFEC_MAINTAIN_ORDER: Maintain order execution of blocks. + * @XSDFEC_OUT_OF_ORDER: Out-of-order execution of blocks. + * + * This enum is used to indicate whether the order of blocks can change from + * input to output. + */ +enum xsdfec_order { + XSDFEC_MAINTAIN_ORDER = 0, + XSDFEC_OUT_OF_ORDER, +}; + +/** + * enum xsdfec_state - State. + * @XSDFEC_INIT: Driver is initialized. + * @XSDFEC_STARTED: Driver is started. + * @XSDFEC_STOPPED: Driver is stopped. + * @XSDFEC_NEEDS_RESET: Driver needs to be reset. + * @XSDFEC_PL_RECONFIGURE: Programmable Logic needs to be recofigured. + * + * This enum is used to indicate the state of the driver. + */ +enum xsdfec_state { + XSDFEC_INIT = 0, + XSDFEC_STARTED, + XSDFEC_STOPPED, + XSDFEC_NEEDS_RESET, + XSDFEC_PL_RECONFIGURE, +}; + +/** + * enum xsdfec_axis_width - AXIS_WIDTH.DIN Setting for 128-bit width. + * @XSDFEC_1x128b: DIN data input stream consists of a 128-bit lane + * @XSDFEC_2x128b: DIN data input stream consists of two 128-bit lanes + * @XSDFEC_4x128b: DIN data input stream consists of four 128-bit lanes + * + * This enum is used to indicate the AXIS_WIDTH.DIN setting for 128-bit width. + * The number of lanes of the DIN data input stream depends upon the + * AXIS_WIDTH.DIN parameter. + */ +enum xsdfec_axis_width { + XSDFEC_1x128b = 1, + XSDFEC_2x128b = 2, + XSDFEC_4x128b = 4, +}; + +/** + * enum xsdfec_axis_word_include - Words Configuration. + * @XSDFEC_FIXED_VALUE: Fixed, the DIN_WORDS AXI4-Stream interface is removed + * from the IP instance and is driven with the specified + * number of words. + * @XSDFEC_IN_BLOCK: In Block, configures the IP instance to expect a single + * DIN_WORDS value per input code block. The DIN_WORDS + * interface is present. + * @XSDFEC_PER_AXI_TRANSACTION: Per Transaction, configures the IP instance to + * expect one DIN_WORDS value per input transaction on the DIN interface. The + * DIN_WORDS interface is present. + * @XSDFEC_AXIS_WORDS_INCLUDE_MAX: Used to indicate out of bound Words + * Configurations. + * + * This enum is used to specify the DIN_WORDS configuration. + */ +enum xsdfec_axis_word_include { + XSDFEC_FIXED_VALUE = 0, + XSDFEC_IN_BLOCK, + XSDFEC_PER_AXI_TRANSACTION, + XSDFEC_AXIS_WORDS_INCLUDE_MAX, +}; + +/** + * struct xsdfec_irq - Enabling or Disabling Interrupts. + * @enable_isr: If true enables the ISR + * @enable_ecc_isr: If true enables the ECC ISR + */ +struct xsdfec_irq { + __s8 enable_isr; + __s8 enable_ecc_isr; +}; + +/** + * struct xsdfec_config - Configuration of SD-FEC core. + * @code: The codes being used by the SD-FEC instance + * @order: Order of Operation + * @din_width: Width of the DIN AXI4-Stream + * @din_word_include: How DIN_WORDS are inputted + * @dout_width: Width of the DOUT AXI4-Stream + * @dout_word_include: HOW DOUT_WORDS are outputted + * @irq: Enabling or disabling interrupts + * @bypass: Is the core being bypassed + * @code_wr_protect: Is write protection of LDPC codes enabled + */ +struct xsdfec_config { + __u32 code; + __u32 order; + __u32 din_width; + __u32 din_word_include; + __u32 dout_width; + __u32 dout_word_include; + struct xsdfec_irq irq; + __s8 bypass; + __s8 code_wr_protect; +}; + +/* + * XSDFEC IOCTL List + */ +#define XSDFEC_MAGIC 'f' +#endif /* __XILINX_SDFEC_H__ */ -- cgit v1.2.3-59-g8ed1b From 6f86ed820178ba35f87712548e6cd43b91608a6c Mon Sep 17 00:00:00 2001 From: Dragan Cvetic Date: Sat, 27 Jul 2019 09:33:52 +0100 Subject: misc: xilinx_sdfec: Add ability to configure turbo Add the capability to configure and retrieve turbo mode via the ioctls XSDFEC_SET_TURBO and XSDFEC_GET_TURBO. Add char device interface per DT node present and support file operations: - open(), - close(), - unlocked_ioctl(), - compat_ioctl(). Tested-by: Dragan Cvetic Signed-off-by: Derek Kiernan Signed-off-by: Dragan Cvetic Link: https://lore.kernel.org/r/1564216438-322406-3-git-send-email-dragan.cvetic@xilinx.com Signed-off-by: Greg Kroah-Hartman --- drivers/misc/xilinx_sdfec.c | 110 +++++++++++++++++++++++++++++++++++++++ include/uapi/misc/xilinx_sdfec.h | 67 ++++++++++++++++++++++++ 2 files changed, 177 insertions(+) (limited to 'include') diff --git a/drivers/misc/xilinx_sdfec.c b/drivers/misc/xilinx_sdfec.c index 24d9f79fe073..d3dba7e6f896 100644 --- a/drivers/misc/xilinx_sdfec.c +++ b/drivers/misc/xilinx_sdfec.c @@ -19,6 +19,7 @@ #include #include #include +#include #include @@ -106,6 +107,12 @@ static struct mutex dev_idr_lock; /* BYPASS Register */ #define XSDFEC_BYPASS_ADDR (0x3C) +/* Turbo Code Register */ +#define XSDFEC_TURBO_ADDR (0x100) +#define XSDFEC_TURBO_SCALE_MASK (0xFFF) +#define XSDFEC_TURBO_SCALE_BIT_POS (8) +#define XSDFEC_TURBO_SCALE_MAX (15) + /** * struct xsdfec_clks - For managing SD-FEC clocks * @core_clk: Main processing clock for core @@ -214,6 +221,55 @@ static void update_config_from_hw(struct xsdfec_dev *xsdfec) xsdfec->state = XSDFEC_STOPPED; } +static int xsdfec_set_turbo(struct xsdfec_dev *xsdfec, void __user *arg) +{ + struct xsdfec_turbo turbo; + int err; + u32 turbo_write; + + err = copy_from_user(&turbo, arg, sizeof(turbo)); + if (err) + return -EFAULT; + + if (turbo.alg >= XSDFEC_TURBO_ALG_MAX) + return -EINVAL; + + if (turbo.scale > XSDFEC_TURBO_SCALE_MAX) + return -EINVAL; + + /* Check to see what device tree says about the FEC codes */ + if (xsdfec->config.code == XSDFEC_LDPC_CODE) + return -EIO; + + turbo_write = ((turbo.scale & XSDFEC_TURBO_SCALE_MASK) + << XSDFEC_TURBO_SCALE_BIT_POS) | + turbo.alg; + xsdfec_regwrite(xsdfec, XSDFEC_TURBO_ADDR, turbo_write); + return err; +} + +static int xsdfec_get_turbo(struct xsdfec_dev *xsdfec, void __user *arg) +{ + u32 reg_value; + struct xsdfec_turbo turbo_params; + int err; + + if (xsdfec->config.code == XSDFEC_LDPC_CODE) + return -EIO; + + reg_value = xsdfec_regread(xsdfec, XSDFEC_TURBO_ADDR); + + turbo_params.scale = (reg_value & XSDFEC_TURBO_SCALE_MASK) >> + XSDFEC_TURBO_SCALE_BIT_POS; + turbo_params.alg = reg_value & 0x1; + + err = copy_to_user(arg, &turbo_params, sizeof(turbo_params)); + if (err) + err = -EFAULT; + + return err; +} + static u32 xsdfec_translate_axis_width_cfg_val(enum xsdfec_axis_width axis_width_cfg) { @@ -277,8 +333,62 @@ static int xsdfec_cfg_axi_streams(struct xsdfec_dev *xsdfec) return 0; } +static int xsdfec_dev_open(struct inode *iptr, struct file *fptr) +{ + return 0; +} + +static int xsdfec_dev_release(struct inode *iptr, struct file *fptr) +{ + return 0; +} + +static long xsdfec_dev_ioctl(struct file *fptr, unsigned int cmd, + unsigned long data) +{ + struct xsdfec_dev *xsdfec; + void __user *arg = NULL; + int rval = -EINVAL; + + xsdfec = container_of(fptr->private_data, struct xsdfec_dev, miscdev); + + /* check if ioctl argument is present and valid */ + if (_IOC_DIR(cmd) != _IOC_NONE) { + arg = (void __user *)data; + if (!arg) + return rval; + } + + switch (cmd) { + case XSDFEC_SET_TURBO: + rval = xsdfec_set_turbo(xsdfec, arg); + break; + case XSDFEC_GET_TURBO: + rval = xsdfec_get_turbo(xsdfec, arg); + break; + default: + /* Should not get here */ + break; + } + return rval; +} + +#ifdef CONFIG_COMPAT +static long xsdfec_dev_compat_ioctl(struct file *file, unsigned int cmd, + unsigned long data) +{ + return xsdfec_dev_ioctl(file, cmd, (unsigned long)compat_ptr(data)); +} +#endif + static const struct file_operations xsdfec_fops = { .owner = THIS_MODULE, + .open = xsdfec_dev_open, + .release = xsdfec_dev_release, + .unlocked_ioctl = xsdfec_dev_ioctl, +#ifdef CONFIG_COMPAT + .compat_ioctl = xsdfec_dev_compat_ioctl, +#endif }; static int xsdfec_parse_of(struct xsdfec_dev *xsdfec) diff --git a/include/uapi/misc/xilinx_sdfec.h b/include/uapi/misc/xilinx_sdfec.h index 330ea25ddfa1..34f32c37261d 100644 --- a/include/uapi/misc/xilinx_sdfec.h +++ b/include/uapi/misc/xilinx_sdfec.h @@ -40,6 +40,22 @@ enum xsdfec_order { XSDFEC_OUT_OF_ORDER, }; +/** + * enum xsdfec_turbo_alg - Turbo Algorithm Type. + * @XSDFEC_MAX_SCALE: Max Log-Map algorithm with extrinsic scaling. When + * scaling is set to this is equivalent to the Max Log-Map + * algorithm. + * @XSDFEC_MAX_STAR: Log-Map algorithm. + * @XSDFEC_TURBO_ALG_MAX: Used to indicate out of bound Turbo algorithms. + * + * This enum specifies which Turbo Decode algorithm is in use. + */ +enum xsdfec_turbo_alg { + XSDFEC_MAX_SCALE = 0, + XSDFEC_MAX_STAR, + XSDFEC_TURBO_ALG_MAX, +}; + /** * enum xsdfec_state - State. * @XSDFEC_INIT: Driver is initialized. @@ -97,6 +113,29 @@ enum xsdfec_axis_word_include { XSDFEC_AXIS_WORDS_INCLUDE_MAX, }; +/** + * struct xsdfec_turbo - User data for Turbo codes. + * @alg: Specifies which Turbo decode algorithm to use + * @scale: Specifies the extrinsic scaling to apply when the Max Scale algorithm + * has been selected + * + * Turbo code structure to communicate parameters to XSDFEC driver. + */ +struct xsdfec_turbo { + __u32 alg; + __u8 scale; +}; + +/** + * struct xsdfec_status - Status of SD-FEC core. + * @state: State of the SD-FEC core + * @activity: Describes if the SD-FEC instance is Active + */ +struct xsdfec_status { + __u32 state; + __s8 activity; +}; + /** * struct xsdfec_irq - Enabling or Disabling Interrupts. * @enable_isr: If true enables the ISR @@ -135,4 +174,32 @@ struct xsdfec_config { * XSDFEC IOCTL List */ #define XSDFEC_MAGIC 'f' +/** + * DOC: XSDFEC_SET_TURBO + * @Parameters + * + * @struct xsdfec_turbo * + * Pointer to the &struct xsdfec_turbo that contains the Turbo decode + * settings for the SD-FEC core + * + * @Description + * + * ioctl that sets the SD-FEC Turbo parameter values + * + * This can only be used when the driver is in the XSDFEC_STOPPED state + */ +#define XSDFEC_SET_TURBO _IOW(XSDFEC_MAGIC, 4, struct xsdfec_turbo) +/** + * DOC: XSDFEC_GET_TURBO + * @Parameters + * + * @struct xsdfec_turbo * + * Pointer to the &struct xsdfec_turbo that contains the current Turbo + * decode settings of the SD-FEC Block + * + * @Description + * + * ioctl that returns SD-FEC turbo param values + */ +#define XSDFEC_GET_TURBO _IOR(XSDFEC_MAGIC, 7, struct xsdfec_turbo) #endif /* __XILINX_SDFEC_H__ */ -- cgit v1.2.3-59-g8ed1b From 20ec628e8007ec75c2f884e00004f39eab6289b5 Mon Sep 17 00:00:00 2001 From: Dragan Cvetic Date: Sat, 27 Jul 2019 09:33:53 +0100 Subject: misc: xilinx_sdfec: Add ability to configure LDPC Add the capability to configure LDPC mode via the ioctl XSDFEC_ADD_LDPC_CODE_PARAMS. Tested-by: Dragan Cvetic Signed-off-by: Derek Kiernan Signed-off-by: Dragan Cvetic Link: https://lore.kernel.org/r/1564216438-322406-4-git-send-email-dragan.cvetic@xilinx.com Signed-off-by: Greg Kroah-Hartman --- drivers/misc/xilinx_sdfec.c | 324 +++++++++++++++++++++++++++++++++++++++ include/uapi/misc/xilinx_sdfec.h | 98 ++++++++++++ 2 files changed, 422 insertions(+) (limited to 'include') diff --git a/drivers/misc/xilinx_sdfec.c b/drivers/misc/xilinx_sdfec.c index d3dba7e6f896..9be4de07eee2 100644 --- a/drivers/misc/xilinx_sdfec.c +++ b/drivers/misc/xilinx_sdfec.c @@ -20,6 +20,7 @@ #include #include #include +#include #include @@ -113,6 +114,57 @@ static struct mutex dev_idr_lock; #define XSDFEC_TURBO_SCALE_BIT_POS (8) #define XSDFEC_TURBO_SCALE_MAX (15) +/* REG0 Register */ +#define XSDFEC_LDPC_CODE_REG0_ADDR_BASE (0x2000) +#define XSDFEC_LDPC_CODE_REG0_ADDR_HIGH (0x27F0) +#define XSDFEC_REG0_N_MIN (4) +#define XSDFEC_REG0_N_MAX (32768) +#define XSDFEC_REG0_N_MUL_P (256) +#define XSDFEC_REG0_N_LSB (0) +#define XSDFEC_REG0_K_MIN (2) +#define XSDFEC_REG0_K_MAX (32766) +#define XSDFEC_REG0_K_MUL_P (256) +#define XSDFEC_REG0_K_LSB (16) + +/* REG1 Register */ +#define XSDFEC_LDPC_CODE_REG1_ADDR_BASE (0x2004) +#define XSDFEC_LDPC_CODE_REG1_ADDR_HIGH (0x27f4) +#define XSDFEC_REG1_PSIZE_MIN (2) +#define XSDFEC_REG1_PSIZE_MAX (512) +#define XSDFEC_REG1_NO_PACKING_MASK (0x400) +#define XSDFEC_REG1_NO_PACKING_LSB (10) +#define XSDFEC_REG1_NM_MASK (0xFF800) +#define XSDFEC_REG1_NM_LSB (11) +#define XSDFEC_REG1_BYPASS_MASK (0x100000) + +/* REG2 Register */ +#define XSDFEC_LDPC_CODE_REG2_ADDR_BASE (0x2008) +#define XSDFEC_LDPC_CODE_REG2_ADDR_HIGH (0x27f8) +#define XSDFEC_REG2_NLAYERS_MIN (1) +#define XSDFEC_REG2_NLAYERS_MAX (256) +#define XSDFEC_REG2_NNMQC_MASK (0xFFE00) +#define XSDFEC_REG2_NMQC_LSB (9) +#define XSDFEC_REG2_NORM_TYPE_MASK (0x100000) +#define XSDFEC_REG2_NORM_TYPE_LSB (20) +#define XSDFEC_REG2_SPECIAL_QC_MASK (0x200000) +#define XSDFEC_REG2_SPEICAL_QC_LSB (21) +#define XSDFEC_REG2_NO_FINAL_PARITY_MASK (0x400000) +#define XSDFEC_REG2_NO_FINAL_PARITY_LSB (22) +#define XSDFEC_REG2_MAX_SCHEDULE_MASK (0x1800000) +#define XSDFEC_REG2_MAX_SCHEDULE_LSB (23) + +/* REG3 Register */ +#define XSDFEC_LDPC_CODE_REG3_ADDR_BASE (0x200C) +#define XSDFEC_LDPC_CODE_REG3_ADDR_HIGH (0x27FC) +#define XSDFEC_REG3_LA_OFF_LSB (8) +#define XSDFEC_REG3_QC_OFF_LSB (16) + +#define XSDFEC_LDPC_REG_JUMP (0x10) +#define XSDFEC_REG_WIDTH_JUMP (4) + +/* The maximum number of pinned pages */ +#define MAX_NUM_PAGES ((XSDFEC_QC_TABLE_DEPTH / PAGE_SIZE) + 1) + /** * struct xsdfec_clks - For managing SD-FEC clocks * @core_clk: Main processing clock for core @@ -270,6 +322,275 @@ static int xsdfec_get_turbo(struct xsdfec_dev *xsdfec, void __user *arg) return err; } +static int xsdfec_reg0_write(struct xsdfec_dev *xsdfec, u32 n, u32 k, u32 psize, + u32 offset) +{ + u32 wdata; + + if (n < XSDFEC_REG0_N_MIN || n > XSDFEC_REG0_N_MAX || + (n > XSDFEC_REG0_N_MUL_P * psize) || n <= k || ((n % psize) != 0)) { + dev_dbg(xsdfec->dev, "N value is not in range"); + return -EINVAL; + } + n <<= XSDFEC_REG0_N_LSB; + + if (k < XSDFEC_REG0_K_MIN || k > XSDFEC_REG0_K_MAX || + (k > XSDFEC_REG0_K_MUL_P * psize) || ((k % psize) != 0)) { + dev_dbg(xsdfec->dev, "K value is not in range"); + return -EINVAL; + } + k = k << XSDFEC_REG0_K_LSB; + wdata = k | n; + + if (XSDFEC_LDPC_CODE_REG0_ADDR_BASE + (offset * XSDFEC_LDPC_REG_JUMP) > + XSDFEC_LDPC_CODE_REG0_ADDR_HIGH) { + dev_dbg(xsdfec->dev, "Writing outside of LDPC reg0 space 0x%x", + XSDFEC_LDPC_CODE_REG0_ADDR_BASE + + (offset * XSDFEC_LDPC_REG_JUMP)); + return -EINVAL; + } + xsdfec_regwrite(xsdfec, + XSDFEC_LDPC_CODE_REG0_ADDR_BASE + + (offset * XSDFEC_LDPC_REG_JUMP), + wdata); + return 0; +} + +static int xsdfec_reg1_write(struct xsdfec_dev *xsdfec, u32 psize, + u32 no_packing, u32 nm, u32 offset) +{ + u32 wdata; + + if (psize < XSDFEC_REG1_PSIZE_MIN || psize > XSDFEC_REG1_PSIZE_MAX) { + dev_dbg(xsdfec->dev, "Psize is not in range"); + return -EINVAL; + } + + if (no_packing != 0 && no_packing != 1) + dev_dbg(xsdfec->dev, "No-packing bit register invalid"); + no_packing = ((no_packing << XSDFEC_REG1_NO_PACKING_LSB) & + XSDFEC_REG1_NO_PACKING_MASK); + + if (nm & ~(XSDFEC_REG1_NM_MASK >> XSDFEC_REG1_NM_LSB)) + dev_dbg(xsdfec->dev, "NM is beyond 10 bits"); + nm = (nm << XSDFEC_REG1_NM_LSB) & XSDFEC_REG1_NM_MASK; + + wdata = nm | no_packing | psize; + if (XSDFEC_LDPC_CODE_REG1_ADDR_BASE + (offset * XSDFEC_LDPC_REG_JUMP) > + XSDFEC_LDPC_CODE_REG1_ADDR_HIGH) { + dev_dbg(xsdfec->dev, "Writing outside of LDPC reg1 space 0x%x", + XSDFEC_LDPC_CODE_REG1_ADDR_BASE + + (offset * XSDFEC_LDPC_REG_JUMP)); + return -EINVAL; + } + xsdfec_regwrite(xsdfec, + XSDFEC_LDPC_CODE_REG1_ADDR_BASE + + (offset * XSDFEC_LDPC_REG_JUMP), + wdata); + return 0; +} + +static int xsdfec_reg2_write(struct xsdfec_dev *xsdfec, u32 nlayers, u32 nmqc, + u32 norm_type, u32 special_qc, u32 no_final_parity, + u32 max_schedule, u32 offset) +{ + u32 wdata; + + if (nlayers < XSDFEC_REG2_NLAYERS_MIN || + nlayers > XSDFEC_REG2_NLAYERS_MAX) { + dev_dbg(xsdfec->dev, "Nlayers is not in range"); + return -EINVAL; + } + + if (nmqc & ~(XSDFEC_REG2_NNMQC_MASK >> XSDFEC_REG2_NMQC_LSB)) + dev_dbg(xsdfec->dev, "NMQC exceeds 11 bits"); + nmqc = (nmqc << XSDFEC_REG2_NMQC_LSB) & XSDFEC_REG2_NNMQC_MASK; + + if (norm_type > 1) + dev_dbg(xsdfec->dev, "Norm type is invalid"); + norm_type = ((norm_type << XSDFEC_REG2_NORM_TYPE_LSB) & + XSDFEC_REG2_NORM_TYPE_MASK); + if (special_qc > 1) + dev_dbg(xsdfec->dev, "Special QC in invalid"); + special_qc = ((special_qc << XSDFEC_REG2_SPEICAL_QC_LSB) & + XSDFEC_REG2_SPECIAL_QC_MASK); + + if (no_final_parity > 1) + dev_dbg(xsdfec->dev, "No final parity check invalid"); + no_final_parity = + ((no_final_parity << XSDFEC_REG2_NO_FINAL_PARITY_LSB) & + XSDFEC_REG2_NO_FINAL_PARITY_MASK); + if (max_schedule & + ~(XSDFEC_REG2_MAX_SCHEDULE_MASK >> XSDFEC_REG2_MAX_SCHEDULE_LSB)) + dev_dbg(xsdfec->dev, "Max Schdule exceeds 2 bits"); + max_schedule = ((max_schedule << XSDFEC_REG2_MAX_SCHEDULE_LSB) & + XSDFEC_REG2_MAX_SCHEDULE_MASK); + + wdata = (max_schedule | no_final_parity | special_qc | norm_type | + nmqc | nlayers); + + if (XSDFEC_LDPC_CODE_REG2_ADDR_BASE + (offset * XSDFEC_LDPC_REG_JUMP) > + XSDFEC_LDPC_CODE_REG2_ADDR_HIGH) { + dev_dbg(xsdfec->dev, "Writing outside of LDPC reg2 space 0x%x", + XSDFEC_LDPC_CODE_REG2_ADDR_BASE + + (offset * XSDFEC_LDPC_REG_JUMP)); + return -EINVAL; + } + xsdfec_regwrite(xsdfec, + XSDFEC_LDPC_CODE_REG2_ADDR_BASE + + (offset * XSDFEC_LDPC_REG_JUMP), + wdata); + return 0; +} + +static int xsdfec_reg3_write(struct xsdfec_dev *xsdfec, u8 sc_off, u8 la_off, + u16 qc_off, u32 offset) +{ + u32 wdata; + + wdata = ((qc_off << XSDFEC_REG3_QC_OFF_LSB) | + (la_off << XSDFEC_REG3_LA_OFF_LSB) | sc_off); + if (XSDFEC_LDPC_CODE_REG3_ADDR_BASE + (offset * XSDFEC_LDPC_REG_JUMP) > + XSDFEC_LDPC_CODE_REG3_ADDR_HIGH) { + dev_dbg(xsdfec->dev, "Writing outside of LDPC reg3 space 0x%x", + XSDFEC_LDPC_CODE_REG3_ADDR_BASE + + (offset * XSDFEC_LDPC_REG_JUMP)); + return -EINVAL; + } + xsdfec_regwrite(xsdfec, + XSDFEC_LDPC_CODE_REG3_ADDR_BASE + + (offset * XSDFEC_LDPC_REG_JUMP), + wdata); + return 0; +} + +static int xsdfec_table_write(struct xsdfec_dev *xsdfec, u32 offset, + u32 *src_ptr, u32 len, const u32 base_addr, + const u32 depth) +{ + u32 reg = 0; + u32 res; + u32 n, i; + u32 *addr = NULL; + struct page *page[MAX_NUM_PAGES]; + + /* + * Writes that go beyond the length of + * Shared Scale(SC) table should fail + */ + if ((XSDFEC_REG_WIDTH_JUMP * (offset + len)) > depth) { + dev_dbg(xsdfec->dev, "Write exceeds SC table length"); + return -EINVAL; + } + + n = (len * XSDFEC_REG_WIDTH_JUMP) / PAGE_SIZE; + if ((len * XSDFEC_REG_WIDTH_JUMP) % PAGE_SIZE) + n += 1; + + res = get_user_pages_fast((unsigned long)src_ptr, n, 0, page); + if (res < n) { + for (i = 0; i < res; i++) + put_page(page[i]); + return -EINVAL; + } + + for (i = 0; i < n; i++) { + addr = kmap(page[i]); + do { + xsdfec_regwrite(xsdfec, + base_addr + ((offset + reg) * + XSDFEC_REG_WIDTH_JUMP), + addr[reg]); + reg++; + } while ((reg < len) && + ((reg * XSDFEC_REG_WIDTH_JUMP) % PAGE_SIZE)); + put_page(page[i]); + } + return reg; +} + +static int xsdfec_add_ldpc(struct xsdfec_dev *xsdfec, void __user *arg) +{ + struct xsdfec_ldpc_params *ldpc; + int ret, n; + + ldpc = kzalloc(sizeof(*ldpc), GFP_KERNEL); + if (!ldpc) + return -ENOMEM; + + ret = copy_from_user(ldpc, arg, sizeof(*ldpc)); + if (ret) + goto err_out; + + if (xsdfec->config.code == XSDFEC_TURBO_CODE) { + ret = -EIO; + goto err_out; + } + + /* Verify Device has not started */ + if (xsdfec->state == XSDFEC_STARTED) { + ret = -EIO; + goto err_out; + } + + if (xsdfec->config.code_wr_protect) { + ret = -EIO; + goto err_out; + } + + /* Write Reg 0 */ + ret = xsdfec_reg0_write(xsdfec, ldpc->n, ldpc->k, ldpc->psize, + ldpc->code_id); + if (ret) + goto err_out; + + /* Write Reg 1 */ + ret = xsdfec_reg1_write(xsdfec, ldpc->psize, ldpc->no_packing, ldpc->nm, + ldpc->code_id); + if (ret) + goto err_out; + + /* Write Reg 2 */ + ret = xsdfec_reg2_write(xsdfec, ldpc->nlayers, ldpc->nmqc, + ldpc->norm_type, ldpc->special_qc, + ldpc->no_final_parity, ldpc->max_schedule, + ldpc->code_id); + if (ret) + goto err_out; + + /* Write Reg 3 */ + ret = xsdfec_reg3_write(xsdfec, ldpc->sc_off, ldpc->la_off, + ldpc->qc_off, ldpc->code_id); + if (ret) + goto err_out; + + /* Write Shared Codes */ + n = ldpc->nlayers / 4; + if (ldpc->nlayers % 4) + n++; + + ret = xsdfec_table_write(xsdfec, ldpc->sc_off, ldpc->sc_table, n, + XSDFEC_LDPC_SC_TABLE_ADDR_BASE, + XSDFEC_SC_TABLE_DEPTH); + if (ret < 0) + goto err_out; + + ret = xsdfec_table_write(xsdfec, 4 * ldpc->la_off, ldpc->la_table, + ldpc->nlayers, XSDFEC_LDPC_LA_TABLE_ADDR_BASE, + XSDFEC_LA_TABLE_DEPTH); + if (ret < 0) + goto err_out; + + ret = xsdfec_table_write(xsdfec, 4 * ldpc->qc_off, ldpc->qc_table, + ldpc->nqc, XSDFEC_LDPC_QC_TABLE_ADDR_BASE, + XSDFEC_QC_TABLE_DEPTH); + if (ret > 0) + ret = 0; +err_out: + kfree(ldpc); + return ret; +} + static u32 xsdfec_translate_axis_width_cfg_val(enum xsdfec_axis_width axis_width_cfg) { @@ -366,6 +687,9 @@ static long xsdfec_dev_ioctl(struct file *fptr, unsigned int cmd, case XSDFEC_GET_TURBO: rval = xsdfec_get_turbo(xsdfec, arg); break; + case XSDFEC_ADD_LDPC_CODE_PARAMS: + rval = xsdfec_add_ldpc(xsdfec, arg); + break; default: /* Should not get here */ break; diff --git a/include/uapi/misc/xilinx_sdfec.h b/include/uapi/misc/xilinx_sdfec.h index 34f32c37261d..ce532a5e3e81 100644 --- a/include/uapi/misc/xilinx_sdfec.h +++ b/include/uapi/misc/xilinx_sdfec.h @@ -13,6 +13,22 @@ #include +/* Shared LDPC Tables */ +#define XSDFEC_LDPC_SC_TABLE_ADDR_BASE (0x10000) +#define XSDFEC_LDPC_SC_TABLE_ADDR_HIGH (0x10400) +#define XSDFEC_LDPC_LA_TABLE_ADDR_BASE (0x18000) +#define XSDFEC_LDPC_LA_TABLE_ADDR_HIGH (0x19000) +#define XSDFEC_LDPC_QC_TABLE_ADDR_BASE (0x20000) +#define XSDFEC_LDPC_QC_TABLE_ADDR_HIGH (0x28000) + +/* LDPC tables depth */ +#define XSDFEC_SC_TABLE_DEPTH \ + (XSDFEC_LDPC_SC_TABLE_ADDR_HIGH - XSDFEC_LDPC_SC_TABLE_ADDR_BASE) +#define XSDFEC_LA_TABLE_DEPTH \ + (XSDFEC_LDPC_LA_TABLE_ADDR_HIGH - XSDFEC_LDPC_LA_TABLE_ADDR_BASE) +#define XSDFEC_QC_TABLE_DEPTH \ + (XSDFEC_LDPC_QC_TABLE_ADDR_HIGH - XSDFEC_LDPC_QC_TABLE_ADDR_BASE) + /** * enum xsdfec_code - Code Type. * @XSDFEC_TURBO_CODE: Driver is configured for Turbo mode. @@ -126,6 +142,53 @@ struct xsdfec_turbo { __u8 scale; }; +/** + * struct xsdfec_ldpc_params - User data for LDPC codes. + * @n: Number of code word bits + * @k: Number of information bits + * @psize: Size of sub-matrix + * @nlayers: Number of layers in code + * @nqc: Quasi Cyclic Number + * @nmqc: Number of M-sized QC operations in parity check matrix + * @nm: Number of M-size vectors in N + * @norm_type: Normalization required or not + * @no_packing: Determines if multiple QC ops should be performed + * @special_qc: Sub-Matrix property for Circulant weight > 0 + * @no_final_parity: Decide if final parity check needs to be performed + * @max_schedule: Experimental code word scheduling limit + * @sc_off: SC offset + * @la_off: LA offset + * @qc_off: QC offset + * @sc_table: Pointer to SC Table which must be page aligned + * @la_table: Pointer to LA Table which must be page aligned + * @qc_table: Pointer to QC Table which must be page aligned + * @code_id: LDPC Code + * + * This structure describes the LDPC code that is passed to the driver by the + * application. + */ +struct xsdfec_ldpc_params { + __u32 n; + __u32 k; + __u32 psize; + __u32 nlayers; + __u32 nqc; + __u32 nmqc; + __u32 nm; + __u32 norm_type; + __u32 no_packing; + __u32 special_qc; + __u32 no_final_parity; + __u32 max_schedule; + __u32 sc_off; + __u32 la_off; + __u32 qc_off; + __u32 *sc_table; + __u32 *la_table; + __u32 *qc_table; + __u16 code_id; +}; + /** * struct xsdfec_status - Status of SD-FEC core. * @state: State of the SD-FEC core @@ -170,6 +233,20 @@ struct xsdfec_config { __s8 code_wr_protect; }; +/** + * struct xsdfec_ldpc_param_table_sizes - Used to store sizes of SD-FEC table + * entries for an individual LPDC code + * parameter. + * @sc_size: Size of SC table used + * @la_size: Size of LA table used + * @qc_size: Size of QC table used + */ +struct xsdfec_ldpc_param_table_sizes { + __u32 sc_size; + __u32 la_size; + __u32 qc_size; +}; + /* * XSDFEC IOCTL List */ @@ -189,6 +266,27 @@ struct xsdfec_config { * This can only be used when the driver is in the XSDFEC_STOPPED state */ #define XSDFEC_SET_TURBO _IOW(XSDFEC_MAGIC, 4, struct xsdfec_turbo) +/** + * DOC: XSDFEC_ADD_LDPC_CODE_PARAMS + * @Parameters + * + * @struct xsdfec_ldpc_params * + * Pointer to the &struct xsdfec_ldpc_params that contains the LDPC code + * parameters to be added to the SD-FEC Block + * + * @Description + * ioctl to add an LDPC code to the SD-FEC LDPC codes + * + * This can only be used when: + * + * - Driver is in the XSDFEC_STOPPED state + * + * - SD-FEC core is configured as LPDC + * + * - SD-FEC Code Write Protection is disabled + */ +#define XSDFEC_ADD_LDPC_CODE_PARAMS \ + _IOW(XSDFEC_MAGIC, 5, struct xsdfec_ldpc_params) /** * DOC: XSDFEC_GET_TURBO * @Parameters -- cgit v1.2.3-59-g8ed1b From 77dd39d924e650cd20696d790f861dfe26e0cb64 Mon Sep 17 00:00:00 2001 From: Dragan Cvetic Date: Sat, 27 Jul 2019 09:33:54 +0100 Subject: misc: xilinx_sdfec: Add ability to get/set config - Add capability to get SD-FEC config data using ioctl XSDFEC_GET_CONFIG. - Add capability to set SD-FEC data order using ioctl SDFEC_SET_ORDER. - Add capability to set SD-FEC bypass option using ioctl XSDFEC_SET_BYPASS. - Add capability to set SD-FEC active state using ioctl XSDFEC_IS_ACTIVE. Tested-by: Dragan Cvetic Signed-off-by: Derek Kiernan Signed-off-by: Dragan Cvetic Link: https://lore.kernel.org/r/1564216438-322406-5-git-send-email-dragan.cvetic@xilinx.com Signed-off-by: Greg Kroah-Hartman --- drivers/misc/xilinx_sdfec.c | 88 ++++++++++++++++++++++++++++++++++++++++ include/uapi/misc/xilinx_sdfec.h | 57 ++++++++++++++++++++++++++ 2 files changed, 145 insertions(+) (limited to 'include') diff --git a/drivers/misc/xilinx_sdfec.c b/drivers/misc/xilinx_sdfec.c index 9be4de07eee2..579f23617dd7 100644 --- a/drivers/misc/xilinx_sdfec.c +++ b/drivers/misc/xilinx_sdfec.c @@ -273,6 +273,17 @@ static void update_config_from_hw(struct xsdfec_dev *xsdfec) xsdfec->state = XSDFEC_STOPPED; } +static int xsdfec_get_config(struct xsdfec_dev *xsdfec, void __user *arg) +{ + int err; + + err = copy_to_user(arg, &xsdfec->config, sizeof(xsdfec->config)); + if (err) + err = -EFAULT; + + return err; +} + static int xsdfec_set_turbo(struct xsdfec_dev *xsdfec, void __user *arg) { struct xsdfec_turbo turbo; @@ -591,6 +602,71 @@ err_out: return ret; } +static int xsdfec_set_order(struct xsdfec_dev *xsdfec, void __user *arg) +{ + bool order_invalid; + enum xsdfec_order order; + int err; + + err = get_user(order, (enum xsdfec_order *)arg); + if (err) + return -EFAULT; + + order_invalid = (order != XSDFEC_MAINTAIN_ORDER) && + (order != XSDFEC_OUT_OF_ORDER); + if (order_invalid) + return -EINVAL; + + /* Verify Device has not started */ + if (xsdfec->state == XSDFEC_STARTED) + return -EIO; + + xsdfec_regwrite(xsdfec, XSDFEC_ORDER_ADDR, order); + + xsdfec->config.order = order; + + return 0; +} + +static int xsdfec_set_bypass(struct xsdfec_dev *xsdfec, bool __user *arg) +{ + bool bypass; + int err; + + err = get_user(bypass, arg); + if (err) + return -EFAULT; + + /* Verify Device has not started */ + if (xsdfec->state == XSDFEC_STARTED) + return -EIO; + + if (bypass) + xsdfec_regwrite(xsdfec, XSDFEC_BYPASS_ADDR, 1); + else + xsdfec_regwrite(xsdfec, XSDFEC_BYPASS_ADDR, 0); + + xsdfec->config.bypass = bypass; + + return 0; +} + +static int xsdfec_is_active(struct xsdfec_dev *xsdfec, bool __user *arg) +{ + u32 reg_value; + bool is_active; + int err; + + reg_value = xsdfec_regread(xsdfec, XSDFEC_ACTIVE_ADDR); + /* using a double ! operator instead of casting */ + is_active = !!(reg_value & XSDFEC_IS_ACTIVITY_SET); + err = put_user(is_active, arg); + if (err) + return -EFAULT; + + return err; +} + static u32 xsdfec_translate_axis_width_cfg_val(enum xsdfec_axis_width axis_width_cfg) { @@ -681,6 +757,9 @@ static long xsdfec_dev_ioctl(struct file *fptr, unsigned int cmd, } switch (cmd) { + case XSDFEC_GET_CONFIG: + rval = xsdfec_get_config(xsdfec, arg); + break; case XSDFEC_SET_TURBO: rval = xsdfec_set_turbo(xsdfec, arg); break; @@ -690,6 +769,15 @@ static long xsdfec_dev_ioctl(struct file *fptr, unsigned int cmd, case XSDFEC_ADD_LDPC_CODE_PARAMS: rval = xsdfec_add_ldpc(xsdfec, arg); break; + case XSDFEC_SET_ORDER: + rval = xsdfec_set_order(xsdfec, arg); + break; + case XSDFEC_SET_BYPASS: + rval = xsdfec_set_bypass(xsdfec, arg); + break; + case XSDFEC_IS_ACTIVE: + rval = xsdfec_is_active(xsdfec, (bool __user *)arg); + break; default: /* Should not get here */ break; diff --git a/include/uapi/misc/xilinx_sdfec.h b/include/uapi/misc/xilinx_sdfec.h index ce532a5e3e81..b8897ce41821 100644 --- a/include/uapi/misc/xilinx_sdfec.h +++ b/include/uapi/misc/xilinx_sdfec.h @@ -287,6 +287,19 @@ struct xsdfec_ldpc_param_table_sizes { */ #define XSDFEC_ADD_LDPC_CODE_PARAMS \ _IOW(XSDFEC_MAGIC, 5, struct xsdfec_ldpc_params) +/** + * DOC: XSDFEC_GET_CONFIG + * @Parameters + * + * @struct xsdfec_config * + * Pointer to the &struct xsdfec_config that contains the current + * configuration settings of the SD-FEC Block + * + * @Description + * + * ioctl that returns SD-FEC core configuration + */ +#define XSDFEC_GET_CONFIG _IOR(XSDFEC_MAGIC, 6, struct xsdfec_config) /** * DOC: XSDFEC_GET_TURBO * @Parameters @@ -300,4 +313,48 @@ struct xsdfec_ldpc_param_table_sizes { * ioctl that returns SD-FEC turbo param values */ #define XSDFEC_GET_TURBO _IOR(XSDFEC_MAGIC, 7, struct xsdfec_turbo) +/** + * DOC: XSDFEC_SET_ORDER + * @Parameters + * + * @struct unsigned long * + * Pointer to the unsigned long that contains a value from the + * @enum xsdfec_order + * + * @Description + * + * ioctl that sets order, if order of blocks can change from input to output + * + * This can only be used when the driver is in the XSDFEC_STOPPED state + */ +#define XSDFEC_SET_ORDER _IOW(XSDFEC_MAGIC, 8, unsigned long) +/** + * DOC: XSDFEC_SET_BYPASS + * @Parameters + * + * @struct bool * + * Pointer to bool that sets the bypass value, where false results in + * normal operation and false results in the SD-FEC performing the + * configured operations (same number of cycles) but output data matches + * the input data + * + * @Description + * + * ioctl that sets bypass. + * + * This can only be used when the driver is in the XSDFEC_STOPPED state + */ +#define XSDFEC_SET_BYPASS _IOW(XSDFEC_MAGIC, 9, bool) +/** + * DOC: XSDFEC_IS_ACTIVE + * @Parameters + * + * @struct bool * + * Pointer to bool that returns true if the SD-FEC is processing data + * + * @Description + * + * ioctl that determines if SD-FEC is processing data + */ +#define XSDFEC_IS_ACTIVE _IOR(XSDFEC_MAGIC, 10, bool) #endif /* __XILINX_SDFEC_H__ */ -- cgit v1.2.3-59-g8ed1b From cc538f609dee49b73545569c49e3abd891fdd8b3 Mon Sep 17 00:00:00 2001 From: Dragan Cvetic Date: Sat, 27 Jul 2019 09:33:55 +0100 Subject: misc: xilinx_sdfec: Support poll file operation Support monitoring and detecting the SD-FEC error events through IRQ and poll file operation. The SD-FEC device can detect one-error or multi-error events. An error triggers an interrupt which creates and run the ONE_SHOT IRQ thread. The ONE_SHOT IRQ thread detects type of error and pass that information to the poll function. The file_operation callback poll(), collects the events and updates the statistics accordingly. The function poll blocks() on waiting queue which can be unblocked by ONE_SHOT IRQ handling thread. Support SD-FEC interrupt set ioctl callback. The SD-FEC can detect two type of errors: coding errors (ECC) and a data interface errors (TLAST). The errors are events which can trigger an IRQ if enabled. The driver can monitor and detect these errors through IRQ. Also the driver updates the statistical data. Tested-by: Dragan Cvetic Signed-off-by: Derek Kiernan Signed-off-by: Dragan Cvetic Link: https://lore.kernel.org/r/1564216438-322406-6-git-send-email-dragan.cvetic@xilinx.com Signed-off-by: Greg Kroah-Hartman --- drivers/misc/xilinx_sdfec.c | 240 ++++++++++++++++++++++++++++++++++++++- include/uapi/misc/xilinx_sdfec.h | 13 +++ 2 files changed, 249 insertions(+), 4 deletions(-) (limited to 'include') diff --git a/drivers/misc/xilinx_sdfec.c b/drivers/misc/xilinx_sdfec.c index 579f23617dd7..61775c5626ed 100644 --- a/drivers/misc/xilinx_sdfec.c +++ b/drivers/misc/xilinx_sdfec.c @@ -191,27 +191,43 @@ struct xsdfec_clks { * struct xsdfec_dev - Driver data for SDFEC * @miscdev: Misc device handle * @clks: Clocks managed by the SDFEC driver - * @regs: device physical base address - * @dev: pointer to device struct + * @waitq: Driver wait queue * @config: Configuration of the SDFEC device * @dev_name: Device name + * @flags: spinlock flags + * @regs: device physical base address + * @dev: pointer to device struct * @state: State of the SDFEC device * @error_data_lock: Error counter and states spinlock * @dev_id: Device ID + * @isr_err_count: Count of ISR errors + * @cecc_count: Count of Correctable ECC errors (SBE) + * @uecc_count: Count of Uncorrectable ECC errors (MBE) + * @irq: IRQ number + * @state_updated: indicates State updated by interrupt handler + * @stats_updated: indicates Stats updated by interrupt handler * * This structure contains necessary state for SDFEC driver to operate */ struct xsdfec_dev { struct miscdevice miscdev; struct xsdfec_clks clks; - void __iomem *regs; - struct device *dev; + wait_queue_head_t waitq; struct xsdfec_config config; char dev_name[DEV_NAME_LEN]; + unsigned long flags; + void __iomem *regs; + struct device *dev; enum xsdfec_state state; /* Spinlock to protect state_updated and stats_updated */ spinlock_t error_data_lock; int dev_id; + u32 isr_err_count; + u32 cecc_count; + u32 uecc_count; + int irq; + bool state_updated; + bool stats_updated; }; static inline void xsdfec_regwrite(struct xsdfec_dev *xsdfec, u32 addr, @@ -284,6 +300,90 @@ static int xsdfec_get_config(struct xsdfec_dev *xsdfec, void __user *arg) return err; } +static int xsdfec_isr_enable(struct xsdfec_dev *xsdfec, bool enable) +{ + u32 mask_read; + + if (enable) { + /* Enable */ + xsdfec_regwrite(xsdfec, XSDFEC_IER_ADDR, XSDFEC_ISR_MASK); + mask_read = xsdfec_regread(xsdfec, XSDFEC_IMR_ADDR); + if (mask_read & XSDFEC_ISR_MASK) { + dev_dbg(xsdfec->dev, + "SDFEC enabling irq with IER failed"); + return -EIO; + } + } else { + /* Disable */ + xsdfec_regwrite(xsdfec, XSDFEC_IDR_ADDR, XSDFEC_ISR_MASK); + mask_read = xsdfec_regread(xsdfec, XSDFEC_IMR_ADDR); + if ((mask_read & XSDFEC_ISR_MASK) != XSDFEC_ISR_MASK) { + dev_dbg(xsdfec->dev, + "SDFEC disabling irq with IDR failed"); + return -EIO; + } + } + return 0; +} + +static int xsdfec_ecc_isr_enable(struct xsdfec_dev *xsdfec, bool enable) +{ + u32 mask_read; + + if (enable) { + /* Enable */ + xsdfec_regwrite(xsdfec, XSDFEC_ECC_IER_ADDR, + XSDFEC_ALL_ECC_ISR_MASK); + mask_read = xsdfec_regread(xsdfec, XSDFEC_ECC_IMR_ADDR); + if (mask_read & XSDFEC_ALL_ECC_ISR_MASK) { + dev_dbg(xsdfec->dev, + "SDFEC enabling ECC irq with ECC IER failed"); + return -EIO; + } + } else { + /* Disable */ + xsdfec_regwrite(xsdfec, XSDFEC_ECC_IDR_ADDR, + XSDFEC_ALL_ECC_ISR_MASK); + mask_read = xsdfec_regread(xsdfec, XSDFEC_ECC_IMR_ADDR); + if (!(((mask_read & XSDFEC_ALL_ECC_ISR_MASK) == + XSDFEC_ECC_ISR_MASK) || + ((mask_read & XSDFEC_ALL_ECC_ISR_MASK) == + XSDFEC_PL_INIT_ECC_ISR_MASK))) { + dev_dbg(xsdfec->dev, + "SDFEC disable ECC irq with ECC IDR failed"); + return -EIO; + } + } + return 0; +} + +static int xsdfec_set_irq(struct xsdfec_dev *xsdfec, void __user *arg) +{ + struct xsdfec_irq irq; + int err; + int isr_err; + int ecc_err; + + err = copy_from_user(&irq, arg, sizeof(irq)); + if (err) + return -EFAULT; + + /* Setup tlast related IRQ */ + isr_err = xsdfec_isr_enable(xsdfec, irq.enable_isr); + if (!isr_err) + xsdfec->config.irq.enable_isr = irq.enable_isr; + + /* Setup ECC related IRQ */ + ecc_err = xsdfec_ecc_isr_enable(xsdfec, irq.enable_ecc_isr); + if (!ecc_err) + xsdfec->config.irq.enable_ecc_isr = irq.enable_ecc_isr; + + if (isr_err < 0 || ecc_err < 0) + err = -EIO; + + return err; +} + static int xsdfec_set_turbo(struct xsdfec_dev *xsdfec, void __user *arg) { struct xsdfec_turbo turbo; @@ -760,6 +860,9 @@ static long xsdfec_dev_ioctl(struct file *fptr, unsigned int cmd, case XSDFEC_GET_CONFIG: rval = xsdfec_get_config(xsdfec, arg); break; + case XSDFEC_SET_IRQ: + rval = xsdfec_set_irq(xsdfec, arg); + break; case XSDFEC_SET_TURBO: rval = xsdfec_set_turbo(xsdfec, arg); break; @@ -793,11 +896,36 @@ static long xsdfec_dev_compat_ioctl(struct file *file, unsigned int cmd, } #endif +static unsigned int xsdfec_poll(struct file *file, poll_table *wait) +{ + unsigned int mask = 0; + struct xsdfec_dev *xsdfec; + + xsdfec = container_of(file->private_data, struct xsdfec_dev, miscdev); + + if (!xsdfec) + return POLLNVAL | POLLHUP; + + poll_wait(file, &xsdfec->waitq, wait); + + /* XSDFEC ISR detected an error */ + spin_lock_irqsave(&xsdfec->error_data_lock, xsdfec->flags); + if (xsdfec->state_updated) + mask |= POLLIN | POLLPRI; + + if (xsdfec->stats_updated) + mask |= POLLIN | POLLRDNORM; + spin_unlock_irqrestore(&xsdfec->error_data_lock, xsdfec->flags); + + return mask; +} + static const struct file_operations xsdfec_fops = { .owner = THIS_MODULE, .open = xsdfec_dev_open, .release = xsdfec_dev_release, .unlocked_ioctl = xsdfec_dev_ioctl, + .poll = xsdfec_poll, #ifdef CONFIG_COMPAT .compat_ioctl = xsdfec_dev_compat_ioctl, #endif @@ -883,6 +1011,91 @@ static int xsdfec_parse_of(struct xsdfec_dev *xsdfec) return 0; } +static irqreturn_t xsdfec_irq_thread(int irq, void *dev_id) +{ + struct xsdfec_dev *xsdfec = dev_id; + irqreturn_t ret = IRQ_HANDLED; + u32 ecc_err; + u32 isr_err; + u32 uecc_count; + u32 cecc_count; + u32 isr_err_count; + u32 aecc_count; + u32 tmp; + + WARN_ON(xsdfec->irq != irq); + + /* Mask Interrupts */ + xsdfec_isr_enable(xsdfec, false); + xsdfec_ecc_isr_enable(xsdfec, false); + /* Read ISR */ + ecc_err = xsdfec_regread(xsdfec, XSDFEC_ECC_ISR_ADDR); + isr_err = xsdfec_regread(xsdfec, XSDFEC_ISR_ADDR); + /* Clear the interrupts */ + xsdfec_regwrite(xsdfec, XSDFEC_ECC_ISR_ADDR, ecc_err); + xsdfec_regwrite(xsdfec, XSDFEC_ISR_ADDR, isr_err); + + tmp = ecc_err & XSDFEC_ALL_ECC_ISR_MBE_MASK; + /* Count uncorrectable 2-bit errors */ + uecc_count = hweight32(tmp); + /* Count all ECC errors */ + aecc_count = hweight32(ecc_err); + /* Number of correctable 1-bit ECC error */ + cecc_count = aecc_count - 2 * uecc_count; + /* Count ISR errors */ + isr_err_count = hweight32(isr_err); + dev_dbg(xsdfec->dev, "tmp=%x, uecc=%x, aecc=%x, cecc=%x, isr=%x", tmp, + uecc_count, aecc_count, cecc_count, isr_err_count); + dev_dbg(xsdfec->dev, "uecc=%x, cecc=%x, isr=%x", xsdfec->uecc_count, + xsdfec->cecc_count, xsdfec->isr_err_count); + + spin_lock_irqsave(&xsdfec->error_data_lock, xsdfec->flags); + /* Add new errors to a 2-bits counter */ + if (uecc_count) + xsdfec->uecc_count += uecc_count; + /* Add new errors to a 1-bits counter */ + if (cecc_count) + xsdfec->cecc_count += cecc_count; + /* Add new errors to a ISR counter */ + if (isr_err_count) + xsdfec->isr_err_count += isr_err_count; + + /* Update state/stats flag */ + if (uecc_count) { + if (ecc_err & XSDFEC_ECC_ISR_MBE_MASK) + xsdfec->state = XSDFEC_NEEDS_RESET; + else if (ecc_err & XSDFEC_PL_INIT_ECC_ISR_MBE_MASK) + xsdfec->state = XSDFEC_PL_RECONFIGURE; + xsdfec->stats_updated = true; + xsdfec->state_updated = true; + } + + if (cecc_count) + xsdfec->stats_updated = true; + + if (isr_err_count) { + xsdfec->state = XSDFEC_NEEDS_RESET; + xsdfec->stats_updated = true; + xsdfec->state_updated = true; + } + + spin_unlock_irqrestore(&xsdfec->error_data_lock, xsdfec->flags); + dev_dbg(xsdfec->dev, "state=%x, stats=%x", xsdfec->state_updated, + xsdfec->stats_updated); + + /* Enable another polling */ + if (xsdfec->state_updated || xsdfec->stats_updated) + wake_up_interruptible(&xsdfec->waitq); + else + ret = IRQ_NONE; + + /* Unmask Interrupts */ + xsdfec_isr_enable(xsdfec, true); + xsdfec_ecc_isr_enable(xsdfec, true); + + return ret; +} + static int xsdfec_clk_init(struct platform_device *pdev, struct xsdfec_clks *clks) { @@ -1049,6 +1262,7 @@ static int xsdfec_probe(struct platform_device *pdev) struct device *dev; struct resource *res; int err; + bool irq_enabled = true; xsdfec = devm_kzalloc(&pdev->dev, sizeof(*xsdfec), GFP_KERNEL); if (!xsdfec) @@ -1069,6 +1283,12 @@ static int xsdfec_probe(struct platform_device *pdev) goto err_xsdfec_dev; } + xsdfec->irq = platform_get_irq(pdev, 0); + if (xsdfec->irq < 0) { + dev_dbg(dev, "platform_get_irq failed"); + irq_enabled = false; + } + err = xsdfec_parse_of(xsdfec); if (err < 0) goto err_xsdfec_dev; @@ -1078,6 +1298,18 @@ static int xsdfec_probe(struct platform_device *pdev) /* Save driver private data */ platform_set_drvdata(pdev, xsdfec); + if (irq_enabled) { + init_waitqueue_head(&xsdfec->waitq); + /* Register IRQ thread */ + err = devm_request_threaded_irq(dev, xsdfec->irq, NULL, + xsdfec_irq_thread, IRQF_ONESHOT, + "xilinx-sdfec16", xsdfec); + if (err < 0) { + dev_err(dev, "unable to request IRQ%d", xsdfec->irq); + goto err_xsdfec_dev; + } + } + mutex_lock(&dev_idr_lock); err = idr_alloc(&dev_idr, xsdfec->dev_name, 0, 0, GFP_KERNEL); mutex_unlock(&dev_idr_lock); diff --git a/include/uapi/misc/xilinx_sdfec.h b/include/uapi/misc/xilinx_sdfec.h index b8897ce41821..85ee288dbb33 100644 --- a/include/uapi/misc/xilinx_sdfec.h +++ b/include/uapi/misc/xilinx_sdfec.h @@ -251,6 +251,19 @@ struct xsdfec_ldpc_param_table_sizes { * XSDFEC IOCTL List */ #define XSDFEC_MAGIC 'f' +/** + * DOC: XSDFEC_SET_IRQ + * @Parameters + * + * @struct xsdfec_irq * + * Pointer to the &struct xsdfec_irq that contains the interrupt settings + * for the SD-FEC core + * + * @Description + * + * ioctl to enable or disable irq + */ +#define XSDFEC_SET_IRQ _IOW(XSDFEC_MAGIC, 3, struct xsdfec_irq) /** * DOC: XSDFEC_SET_TURBO * @Parameters -- cgit v1.2.3-59-g8ed1b From 6bd6a690c2e7e710aa7ccefa4edc83f14099907e Mon Sep 17 00:00:00 2001 From: Dragan Cvetic Date: Sat, 27 Jul 2019 09:33:56 +0100 Subject: misc: xilinx_sdfec: Add stats & status ioctls SD-FEC statistic data are: - count of data interface errors (isr_err_count) - count of Correctable ECC errors (cecc_count) - count of Uncorrectable ECC errors (uecc_count) Add support: 1. clear stats ioctl callback which clears collected statistic data, 2. get stats ioctl callback which reads a collected statistic data, 3. set default configuration ioctl callback, 4. start ioctl callback enables SD-FEC HW, 5. stop ioctl callback disables SD-FEC HW. In a failed state driver enables the following ioctls: - get status - get statistics - clear stats - set default SD-FEC device configuration Tested-by: Santhosh Dyavanapally Tested by: Punnaiah Choudary Kalluri Tested-by: Derek Kiernan Tested-by: Dragan Cvetic Signed-off-by: Derek Kiernan Signed-off-by: Dragan Cvetic Link: https://lore.kernel.org/r/1564216438-322406-7-git-send-email-dragan.cvetic@xilinx.com Signed-off-by: Greg Kroah-Hartman --- drivers/misc/xilinx_sdfec.c | 125 +++++++++++++++++++++++++++++++++++++++ include/uapi/misc/xilinx_sdfec.h | 75 +++++++++++++++++++++++ 2 files changed, 200 insertions(+) (limited to 'include') diff --git a/drivers/misc/xilinx_sdfec.c b/drivers/misc/xilinx_sdfec.c index 61775c5626ed..45f127ac12f6 100644 --- a/drivers/misc/xilinx_sdfec.c +++ b/drivers/misc/xilinx_sdfec.c @@ -206,6 +206,7 @@ struct xsdfec_clks { * @irq: IRQ number * @state_updated: indicates State updated by interrupt handler * @stats_updated: indicates Stats updated by interrupt handler + * @intr_enabled: indicates IRQ enabled * * This structure contains necessary state for SDFEC driver to operate */ @@ -228,6 +229,7 @@ struct xsdfec_dev { int irq; bool state_updated; bool stats_updated; + bool intr_enabled; }; static inline void xsdfec_regwrite(struct xsdfec_dev *xsdfec, u32 addr, @@ -289,6 +291,25 @@ static void update_config_from_hw(struct xsdfec_dev *xsdfec) xsdfec->state = XSDFEC_STOPPED; } +static int xsdfec_get_status(struct xsdfec_dev *xsdfec, void __user *arg) +{ + struct xsdfec_status status; + int err; + + spin_lock_irqsave(&xsdfec->error_data_lock, xsdfec->flags); + status.state = xsdfec->state; + xsdfec->state_updated = false; + spin_unlock_irqrestore(&xsdfec->error_data_lock, xsdfec->flags); + status.activity = (xsdfec_regread(xsdfec, XSDFEC_ACTIVE_ADDR) & + XSDFEC_IS_ACTIVITY_SET); + + err = copy_to_user(arg, &status, sizeof(status)); + if (err) + err = -EFAULT; + + return err; +} + static int xsdfec_get_config(struct xsdfec_dev *xsdfec, void __user *arg) { int err; @@ -840,6 +861,82 @@ static int xsdfec_dev_release(struct inode *iptr, struct file *fptr) return 0; } +static int xsdfec_start(struct xsdfec_dev *xsdfec) +{ + u32 regread; + + regread = xsdfec_regread(xsdfec, XSDFEC_FEC_CODE_ADDR); + regread &= 0x1; + if (regread != xsdfec->config.code) { + dev_dbg(xsdfec->dev, + "%s SDFEC HW code does not match driver code, reg %d, code %d", + __func__, regread, xsdfec->config.code); + return -EINVAL; + } + + /* Set AXIS enable */ + xsdfec_regwrite(xsdfec, XSDFEC_AXIS_ENABLE_ADDR, + XSDFEC_AXIS_ENABLE_MASK); + /* Done */ + xsdfec->state = XSDFEC_STARTED; + return 0; +} + +static int xsdfec_stop(struct xsdfec_dev *xsdfec) +{ + u32 regread; + + if (xsdfec->state != XSDFEC_STARTED) + dev_dbg(xsdfec->dev, "Device not started correctly"); + /* Disable AXIS_ENABLE Input interfaces only */ + regread = xsdfec_regread(xsdfec, XSDFEC_AXIS_ENABLE_ADDR); + regread &= (~XSDFEC_AXIS_IN_ENABLE_MASK); + xsdfec_regwrite(xsdfec, XSDFEC_AXIS_ENABLE_ADDR, regread); + /* Stop */ + xsdfec->state = XSDFEC_STOPPED; + return 0; +} + +static int xsdfec_clear_stats(struct xsdfec_dev *xsdfec) +{ + spin_lock_irqsave(&xsdfec->error_data_lock, xsdfec->flags); + xsdfec->isr_err_count = 0; + xsdfec->uecc_count = 0; + xsdfec->cecc_count = 0; + spin_unlock_irqrestore(&xsdfec->error_data_lock, xsdfec->flags); + + return 0; +} + +static int xsdfec_get_stats(struct xsdfec_dev *xsdfec, void __user *arg) +{ + int err; + struct xsdfec_stats user_stats; + + spin_lock_irqsave(&xsdfec->error_data_lock, xsdfec->flags); + user_stats.isr_err_count = xsdfec->isr_err_count; + user_stats.cecc_count = xsdfec->cecc_count; + user_stats.uecc_count = xsdfec->uecc_count; + xsdfec->stats_updated = false; + spin_unlock_irqrestore(&xsdfec->error_data_lock, xsdfec->flags); + + err = copy_to_user(arg, &user_stats, sizeof(user_stats)); + if (err) + err = -EFAULT; + + return err; +} + +static int xsdfec_set_default_config(struct xsdfec_dev *xsdfec) +{ + /* Ensure registers are aligned with core configuration */ + xsdfec_regwrite(xsdfec, XSDFEC_FEC_CODE_ADDR, xsdfec->config.code); + xsdfec_cfg_axi_streams(xsdfec); + update_config_from_hw(xsdfec); + + return 0; +} + static long xsdfec_dev_ioctl(struct file *fptr, unsigned int cmd, unsigned long data) { @@ -849,6 +946,16 @@ static long xsdfec_dev_ioctl(struct file *fptr, unsigned int cmd, xsdfec = container_of(fptr->private_data, struct xsdfec_dev, miscdev); + /* In failed state allow only reset and get status IOCTLs */ + if (xsdfec->state == XSDFEC_NEEDS_RESET && + (cmd != XSDFEC_SET_DEFAULT_CONFIG && cmd != XSDFEC_GET_STATUS && + cmd != XSDFEC_GET_STATS && cmd != XSDFEC_CLEAR_STATS)) { + return -EPERM; + } + + if (_IOC_TYPE(cmd) != XSDFEC_MAGIC) + return -ENOTTY; + /* check if ioctl argument is present and valid */ if (_IOC_DIR(cmd) != _IOC_NONE) { arg = (void __user *)data; @@ -857,9 +964,27 @@ static long xsdfec_dev_ioctl(struct file *fptr, unsigned int cmd, } switch (cmd) { + case XSDFEC_START_DEV: + rval = xsdfec_start(xsdfec); + break; + case XSDFEC_STOP_DEV: + rval = xsdfec_stop(xsdfec); + break; + case XSDFEC_CLEAR_STATS: + rval = xsdfec_clear_stats(xsdfec); + break; + case XSDFEC_GET_STATS: + rval = xsdfec_get_stats(xsdfec, arg); + break; + case XSDFEC_GET_STATUS: + rval = xsdfec_get_status(xsdfec, arg); + break; case XSDFEC_GET_CONFIG: rval = xsdfec_get_config(xsdfec, arg); break; + case XSDFEC_SET_DEFAULT_CONFIG: + rval = xsdfec_set_default_config(xsdfec); + break; case XSDFEC_SET_IRQ: rval = xsdfec_set_irq(xsdfec, arg); break; diff --git a/include/uapi/misc/xilinx_sdfec.h b/include/uapi/misc/xilinx_sdfec.h index 85ee288dbb33..ee1a42ae6f46 100644 --- a/include/uapi/misc/xilinx_sdfec.h +++ b/include/uapi/misc/xilinx_sdfec.h @@ -233,6 +233,21 @@ struct xsdfec_config { __s8 code_wr_protect; }; +/** + * struct xsdfec_stats - Stats retrived by ioctl XSDFEC_GET_STATS. Used + * to buffer atomic_t variables from struct + * xsdfec_dev. Counts are accumulated until + * the user clears them. + * @isr_err_count: Count of ISR errors + * @cecc_count: Count of Correctable ECC errors (SBE) + * @uecc_count: Count of Uncorrectable ECC errors (MBE) + */ +struct xsdfec_stats { + __u32 isr_err_count; + __u32 cecc_count; + __u32 uecc_count; +}; + /** * struct xsdfec_ldpc_param_table_sizes - Used to store sizes of SD-FEC table * entries for an individual LPDC code @@ -251,6 +266,32 @@ struct xsdfec_ldpc_param_table_sizes { * XSDFEC IOCTL List */ #define XSDFEC_MAGIC 'f' +/** + * DOC: XSDFEC_START_DEV + * + * @Description + * + * ioctl to start SD-FEC core + * + * This fails if the XSDFEC_SET_ORDER ioctl has not been previously called + */ +#define XSDFEC_START_DEV _IO(XSDFEC_MAGIC, 0) +/** + * DOC: XSDFEC_STOP_DEV + * + * @Description + * + * ioctl to stop the SD-FEC core + */ +#define XSDFEC_STOP_DEV _IO(XSDFEC_MAGIC, 1) +/** + * DOC: XSDFEC_GET_STATUS + * + * @Description + * + * ioctl that returns status of SD-FEC core + */ +#define XSDFEC_GET_STATUS _IOR(XSDFEC_MAGIC, 2, struct xsdfec_status) /** * DOC: XSDFEC_SET_IRQ * @Parameters @@ -370,4 +411,38 @@ struct xsdfec_ldpc_param_table_sizes { * ioctl that determines if SD-FEC is processing data */ #define XSDFEC_IS_ACTIVE _IOR(XSDFEC_MAGIC, 10, bool) +/** + * DOC: XSDFEC_CLEAR_STATS + * + * @Description + * + * ioctl that clears error stats collected during interrupts + */ +#define XSDFEC_CLEAR_STATS _IO(XSDFEC_MAGIC, 11) +/** + * DOC: XSDFEC_GET_STATS + * @Parameters + * + * @struct xsdfec_stats * + * Pointer to the &struct xsdfec_stats that will contain the updated stats + * values + * + * @Description + * + * ioctl that returns SD-FEC core stats + * + * This can only be used when the driver is in the XSDFEC_STOPPED state + */ +#define XSDFEC_GET_STATS _IOR(XSDFEC_MAGIC, 12, struct xsdfec_stats) +/** + * DOC: XSDFEC_SET_DEFAULT_CONFIG + * + * @Description + * + * ioctl that returns SD-FEC core to default config, use after a reset + * + * This can only be used when the driver is in the XSDFEC_STOPPED state + */ +#define XSDFEC_SET_DEFAULT_CONFIG _IO(XSDFEC_MAGIC, 13) + #endif /* __XILINX_SDFEC_H__ */ -- cgit v1.2.3-59-g8ed1b From 6311b6521bcc804e4d2fd45a5640562a7b8b5241 Mon Sep 17 00:00:00 2001 From: Jordan Crouse Date: Mon, 5 Aug 2019 14:33:46 -0600 Subject: drivers: qcom: Add BCM vote macro to header The macro to generate a Bus Controller Manager (BCM) TCS command is used by the interconnect driver but might also be interesting to other drivers that need to construct TCS commands for sub processors so move it out of the sdm845 specific file and into the header. Signed-off-by: Jordan Crouse Acked-by: Stephen Boyd Acked-by: Bjorn Andersson Signed-off-by: Georgi Djakov --- drivers/clk/qcom/clk-rpmh.c | 16 +++------------- drivers/interconnect/qcom/sdm845.c | 19 +------------------ include/soc/qcom/tcs.h | 20 +++++++++++++++++++- 3 files changed, 23 insertions(+), 32 deletions(-) (limited to 'include') diff --git a/drivers/clk/qcom/clk-rpmh.c b/drivers/clk/qcom/clk-rpmh.c index c3fd632af119..a32bfaeb7e61 100644 --- a/drivers/clk/qcom/clk-rpmh.c +++ b/drivers/clk/qcom/clk-rpmh.c @@ -1,6 +1,6 @@ // SPDX-License-Identifier: GPL-2.0 /* - * Copyright (c) 2018, The Linux Foundation. All rights reserved. + * Copyright (c) 2018-2019, The Linux Foundation. All rights reserved. */ #include @@ -12,23 +12,13 @@ #include #include #include +#include #include #define CLK_RPMH_ARC_EN_OFFSET 0 #define CLK_RPMH_VRM_EN_OFFSET 4 -#define BCM_TCS_CMD_COMMIT_MASK 0x40000000 -#define BCM_TCS_CMD_VALID_SHIFT 29 -#define BCM_TCS_CMD_VOTE_MASK 0x3fff -#define BCM_TCS_CMD_VOTE_SHIFT 0 - -#define BCM_TCS_CMD(valid, vote) \ - (BCM_TCS_CMD_COMMIT_MASK | \ - ((valid) << BCM_TCS_CMD_VALID_SHIFT) | \ - ((vote & BCM_TCS_CMD_VOTE_MASK) \ - << BCM_TCS_CMD_VOTE_SHIFT)) - /** * struct bcm_db - Auxiliary data pertaining to each Bus Clock Manager(BCM) * @unit: divisor used to convert Hz value to an RPMh msg @@ -269,7 +259,7 @@ static int clk_rpmh_bcm_send_cmd(struct clk_rpmh *c, bool enable) } cmd.addr = c->res_addr; - cmd.data = BCM_TCS_CMD(enable, cmd_state); + cmd.data = BCM_TCS_CMD(1, enable, 0, cmd_state); ret = rpmh_write_async(c->dev, RPMH_ACTIVE_ONLY_STATE, &cmd, 1); if (ret) { diff --git a/drivers/interconnect/qcom/sdm845.c b/drivers/interconnect/qcom/sdm845.c index 93df67345b39..57955596bb59 100644 --- a/drivers/interconnect/qcom/sdm845.c +++ b/drivers/interconnect/qcom/sdm845.c @@ -1,6 +1,6 @@ // SPDX-License-Identifier: GPL-2.0 /* - * Copyright (c) 2018, The Linux Foundation. All rights reserved. + * Copyright (c) 2018-2019, The Linux Foundation. All rights reserved. * */ @@ -20,23 +20,6 @@ #include #include -#define BCM_TCS_CMD_COMMIT_SHFT 30 -#define BCM_TCS_CMD_COMMIT_MASK 0x40000000 -#define BCM_TCS_CMD_VALID_SHFT 29 -#define BCM_TCS_CMD_VALID_MASK 0x20000000 -#define BCM_TCS_CMD_VOTE_X_SHFT 14 -#define BCM_TCS_CMD_VOTE_MASK 0x3fff -#define BCM_TCS_CMD_VOTE_Y_SHFT 0 -#define BCM_TCS_CMD_VOTE_Y_MASK 0xfffc000 - -#define BCM_TCS_CMD(commit, valid, vote_x, vote_y) \ - (((commit) << BCM_TCS_CMD_COMMIT_SHFT) | \ - ((valid) << BCM_TCS_CMD_VALID_SHFT) | \ - ((cpu_to_le32(vote_x) & \ - BCM_TCS_CMD_VOTE_MASK) << BCM_TCS_CMD_VOTE_X_SHFT) | \ - ((cpu_to_le32(vote_y) & \ - BCM_TCS_CMD_VOTE_MASK) << BCM_TCS_CMD_VOTE_Y_SHFT)) - #define to_qcom_provider(_provider) \ container_of(_provider, struct qcom_icc_provider, provider) diff --git a/include/soc/qcom/tcs.h b/include/soc/qcom/tcs.h index 262876a59e86..7a2a055ba6b0 100644 --- a/include/soc/qcom/tcs.h +++ b/include/soc/qcom/tcs.h @@ -1,6 +1,6 @@ /* SPDX-License-Identifier: GPL-2.0 */ /* - * Copyright (c) 2016-2018, The Linux Foundation. All rights reserved. + * Copyright (c) 2016-2019, The Linux Foundation. All rights reserved. */ #ifndef __SOC_QCOM_TCS_H__ @@ -53,4 +53,22 @@ struct tcs_request { struct tcs_cmd *cmds; }; +#define BCM_TCS_CMD_COMMIT_SHFT 30 +#define BCM_TCS_CMD_COMMIT_MASK 0x40000000 +#define BCM_TCS_CMD_VALID_SHFT 29 +#define BCM_TCS_CMD_VALID_MASK 0x20000000 +#define BCM_TCS_CMD_VOTE_X_SHFT 14 +#define BCM_TCS_CMD_VOTE_MASK 0x3fff +#define BCM_TCS_CMD_VOTE_Y_SHFT 0 +#define BCM_TCS_CMD_VOTE_Y_MASK 0xfffc000 + +/* Construct a Bus Clock Manager (BCM) specific TCS command */ +#define BCM_TCS_CMD(commit, valid, vote_x, vote_y) \ + (((commit) << BCM_TCS_CMD_COMMIT_SHFT) | \ + ((valid) << BCM_TCS_CMD_VALID_SHFT) | \ + ((cpu_to_le32(vote_x) & \ + BCM_TCS_CMD_VOTE_MASK) << BCM_TCS_CMD_VOTE_X_SHFT) | \ + ((cpu_to_le32(vote_y) & \ + BCM_TCS_CMD_VOTE_MASK) << BCM_TCS_CMD_VOTE_Y_SHFT)) + #endif /* __SOC_QCOM_TCS_H__ */ -- cgit v1.2.3-59-g8ed1b From 088e88be5a380cc4e81963a9a02815da465d144f Mon Sep 17 00:00:00 2001 From: Martin Blumenstingl Date: Sat, 27 Jul 2019 14:04:12 +0200 Subject: dt-bindings: phy: add binding for the Lantiq VRX200 and ARX300 PCIe PHYs Add the bindings for the PCIe PHY on Lantiq VRX200 and ARX300 SoCs. The IP block contains settings for the PHY and a PLL. The PLL mode is configurable through a dedicated #phy-cell in .dts. Signed-off-by: Martin Blumenstingl Reviewed-by: Rob Herring Signed-off-by: Kishon Vijay Abraham I --- .../bindings/phy/lantiq,vrx200-pcie-phy.yaml | 95 ++++++++++++++++++++++ include/dt-bindings/phy/phy-lantiq-vrx200-pcie.h | 11 +++ 2 files changed, 106 insertions(+) create mode 100644 Documentation/devicetree/bindings/phy/lantiq,vrx200-pcie-phy.yaml create mode 100644 include/dt-bindings/phy/phy-lantiq-vrx200-pcie.h (limited to 'include') diff --git a/Documentation/devicetree/bindings/phy/lantiq,vrx200-pcie-phy.yaml b/Documentation/devicetree/bindings/phy/lantiq,vrx200-pcie-phy.yaml new file mode 100644 index 000000000000..8a56a8526cef --- /dev/null +++ b/Documentation/devicetree/bindings/phy/lantiq,vrx200-pcie-phy.yaml @@ -0,0 +1,95 @@ +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) +%YAML 1.2 +--- +$id: http://devicetree.org/schemas/phy/lantiq,vrx200-pcie-phy.yaml# +$schema: http://devicetree.org/meta-schemas/core.yaml# + +title: Lantiq VRX200 and ARX300 PCIe PHY Device Tree Bindings + +maintainers: + - Martin Blumenstingl + +properties: + "#phy-cells": + const: 1 + description: selects the PHY mode as defined in + + compatible: + enum: + - lantiq,vrx200-pcie-phy + - lantiq,arx300-pcie-phy + + reg: + maxItems: 1 + + clocks: + items: + - description: PHY module clock + - description: PDI register clock + + clock-names: + items: + - const: phy + - const: pdi + + resets: + items: + - description: exclusive PHY reset line + - description: shared reset line between the PCIe PHY and PCIe controller + + resets-names: + items: + - const: phy + - const: pcie + + lantiq,rcu: + $ref: /schemas/types.yaml#/definitions/phandle + description: phandle to the RCU syscon + + lantiq,rcu-endian-offset: + $ref: /schemas/types.yaml#/definitions/uint32 + description: the offset of the endian registers for this PHY instance in the RCU syscon + + lantiq,rcu-big-endian-mask: + $ref: /schemas/types.yaml#/definitions/uint32 + description: the mask to set the PDI (PHY) registers for this PHY instance to big endian + + big-endian: + description: Configures the PDI (PHY) registers in big-endian mode + type: boolean + + little-endian: + description: Configures the PDI (PHY) registers in big-endian mode + type: boolean + +required: + - "#phy-cells" + - compatible + - reg + - clocks + - clock-names + - resets + - reset-names + - lantiq,rcu + - lantiq,rcu-endian-offset + - lantiq,rcu-big-endian-mask + +additionalProperties: false + +examples: + - | + pcie0_phy: phy@106800 { + compatible = "lantiq,vrx200-pcie-phy"; + reg = <0x106800 0x100>; + lantiq,rcu = <&rcu0>; + lantiq,rcu-endian-offset = <0x4c>; + lantiq,rcu-big-endian-mask = <0x80>; /* bit 7 */ + big-endian; + clocks = <&pmu 32>, <&pmu 36>; + clock-names = "phy", "pdi"; + resets = <&reset0 12 24>, <&reset0 22 22>; + reset-names = "phy", "pcie"; + #phy-cells = <1>; + }; + +... diff --git a/include/dt-bindings/phy/phy-lantiq-vrx200-pcie.h b/include/dt-bindings/phy/phy-lantiq-vrx200-pcie.h new file mode 100644 index 000000000000..95a7896356d6 --- /dev/null +++ b/include/dt-bindings/phy/phy-lantiq-vrx200-pcie.h @@ -0,0 +1,11 @@ +/* SPDX-License-Identifier: GPL-2.0-only */ +/* + * Copyright (C) 2019 Martin Blumenstingl + */ + +#define LANTIQ_PCIE_PHY_MODE_25MHZ 0 +#define LANTIQ_PCIE_PHY_MODE_25MHZ_SSC 1 +#define LANTIQ_PCIE_PHY_MODE_36MHZ 2 +#define LANTIQ_PCIE_PHY_MODE_36MHZ_SSC 3 +#define LANTIQ_PCIE_PHY_MODE_100MHZ 4 +#define LANTIQ_PCIE_PHY_MODE_100MHZ_SSC 5 -- cgit v1.2.3-59-g8ed1b From 3cdb9446a117d5d63af823bde6fe6babc312e77b Mon Sep 17 00:00:00 2001 From: Mika Westerberg Date: Tue, 16 Jan 2018 22:19:00 +0200 Subject: thunderbolt: Add support for Intel Ice Lake The Thunderbolt controller is integrated into the Ice Lake CPU itself and requires special flows to power it on and off using force power bit in NHI VSEC registers. Runtime PM (RTD3) and Sx flows also differ from the discrete solutions. Now the firmware notifies the driver whether RTD3 entry or exit are possible. The driver is responsible of sending Go2Sx command through link controller mailbox when system enters Sx states (suspend-to-mem/disk). Rest of the ICM firwmare flows follow Titan Ridge. Signed-off-by: Raanan Avargil Signed-off-by: Mika Westerberg Reviewed-by: Yehezkel Bernat Tested-by: Mario Limonciello --- drivers/thunderbolt/Makefile | 2 +- drivers/thunderbolt/icm.c | 176 +++++++++++++++++++++++++++++++++++++--- drivers/thunderbolt/nhi.c | 112 ++++++++++++++++++++++++-- drivers/thunderbolt/nhi.h | 22 +++++ drivers/thunderbolt/nhi_ops.c | 179 +++++++++++++++++++++++++++++++++++++++++ drivers/thunderbolt/nhi_regs.h | 37 +++++++++ drivers/thunderbolt/switch.c | 2 + drivers/thunderbolt/tb_msgs.h | 16 +++- include/linux/thunderbolt.h | 2 + 9 files changed, 526 insertions(+), 22 deletions(-) create mode 100644 drivers/thunderbolt/nhi_ops.c (limited to 'include') diff --git a/drivers/thunderbolt/Makefile b/drivers/thunderbolt/Makefile index 3f55cb3c81b2..001187c577bf 100644 --- a/drivers/thunderbolt/Makefile +++ b/drivers/thunderbolt/Makefile @@ -1,4 +1,4 @@ # SPDX-License-Identifier: GPL-2.0-only obj-${CONFIG_THUNDERBOLT} := thunderbolt.o -thunderbolt-objs := nhi.o ctl.o tb.o switch.o cap.o path.o tunnel.o eeprom.o +thunderbolt-objs := nhi.o nhi_ops.o ctl.o tb.o switch.o cap.o path.o tunnel.o eeprom.o thunderbolt-objs += domain.o dma_port.o icm.o property.o xdomain.o lc.o diff --git a/drivers/thunderbolt/icm.c b/drivers/thunderbolt/icm.c index e9835ab35465..245588f691e7 100644 --- a/drivers/thunderbolt/icm.c +++ b/drivers/thunderbolt/icm.c @@ -56,16 +56,19 @@ * @max_boot_acl: Maximum number of preboot ACL entries (%0 if not supported) * @rpm: Does the controller support runtime PM (RTD3) * @can_upgrade_nvm: Can the NVM firmware be upgrade on this controller + * @veto: Is RTD3 veto in effect * @is_supported: Checks if we can support ICM on this controller * @cio_reset: Trigger CIO reset * @get_mode: Read and return the ICM firmware mode (optional) * @get_route: Find a route string for given switch * @save_devices: Ask ICM to save devices to ACL when suspending (optional) * @driver_ready: Send driver ready message to ICM + * @set_uuid: Set UUID for the root switch (optional) * @device_connected: Handle device connected ICM message * @device_disconnected: Handle device disconnected ICM message * @xdomain_connected - Handle XDomain connected ICM message * @xdomain_disconnected - Handle XDomain disconnected ICM message + * @rtd3_veto: Handle RTD3 veto notification ICM message */ struct icm { struct mutex request_lock; @@ -76,6 +79,7 @@ struct icm { bool safe_mode; bool rpm; bool can_upgrade_nvm; + bool veto; bool (*is_supported)(struct tb *tb); int (*cio_reset)(struct tb *tb); int (*get_mode)(struct tb *tb); @@ -84,6 +88,7 @@ struct icm { int (*driver_ready)(struct tb *tb, enum tb_security_level *security_level, size_t *nboot_acl, bool *rpm); + void (*set_uuid)(struct tb *tb); void (*device_connected)(struct tb *tb, const struct icm_pkg_header *hdr); void (*device_disconnected)(struct tb *tb, @@ -92,6 +97,7 @@ struct icm { const struct icm_pkg_header *hdr); void (*xdomain_disconnected)(struct tb *tb, const struct icm_pkg_header *hdr); + void (*rtd3_veto)(struct tb *tb, const struct icm_pkg_header *hdr); }; struct icm_notification { @@ -296,6 +302,43 @@ static int icm_request(struct tb *tb, const void *request, size_t request_size, return -ETIMEDOUT; } +/* + * If rescan is queued to run (we are resuming), postpone it to give the + * firmware some more time to send device connected notifications for next + * devices in the chain. + */ +static void icm_postpone_rescan(struct tb *tb) +{ + struct icm *icm = tb_priv(tb); + + if (delayed_work_pending(&icm->rescan_work)) + mod_delayed_work(tb->wq, &icm->rescan_work, + msecs_to_jiffies(500)); +} + +static void icm_veto_begin(struct tb *tb) +{ + struct icm *icm = tb_priv(tb); + + if (!icm->veto) { + icm->veto = true; + /* Keep the domain powered while veto is in effect */ + pm_runtime_get(&tb->dev); + } +} + +static void icm_veto_end(struct tb *tb) +{ + struct icm *icm = tb_priv(tb); + + if (icm->veto) { + icm->veto = false; + /* Allow the domain suspend now */ + pm_runtime_mark_last_busy(&tb->dev); + pm_runtime_put_autosuspend(&tb->dev); + } +} + static bool icm_fr_is_supported(struct tb *tb) { return !x86_apple_machine; @@ -519,14 +562,16 @@ static int icm_fr_disconnect_xdomain_paths(struct tb *tb, struct tb_xdomain *xd) return 0; } -static void add_switch(struct tb_switch *parent_sw, u64 route, - const uuid_t *uuid, const u8 *ep_name, - size_t ep_name_size, u8 connection_id, u8 connection_key, - u8 link, u8 depth, enum tb_security_level security_level, - bool authorized, bool boot) +static struct tb_switch *add_switch(struct tb_switch *parent_sw, u64 route, + const uuid_t *uuid, const u8 *ep_name, + size_t ep_name_size, u8 connection_id, + u8 connection_key, u8 link, u8 depth, + enum tb_security_level security_level, + bool authorized, bool boot) { const struct intel_vss *vss; struct tb_switch *sw; + int ret; pm_runtime_get_sync(&parent_sw->dev); @@ -557,14 +602,18 @@ static void add_switch(struct tb_switch *parent_sw, u64 route, tb_port_at(route, parent_sw)->remote = tb_upstream_port(sw); tb_upstream_port(sw)->remote = tb_port_at(route, parent_sw); - if (tb_switch_add(sw)) { + ret = tb_switch_add(sw); + if (ret) { tb_port_at(tb_route(sw), parent_sw)->remote = NULL; tb_switch_put(sw); + sw = ERR_PTR(ret); } out: pm_runtime_mark_last_busy(&parent_sw->dev); pm_runtime_put_autosuspend(&parent_sw->dev); + + return sw; } static void update_switch(struct tb_switch *parent_sw, struct tb_switch *sw, @@ -656,6 +705,8 @@ icm_fr_device_connected(struct tb *tb, const struct icm_pkg_header *hdr) u64 route; int ret; + icm_postpone_rescan(tb); + link = pkg->link_info & ICM_LINK_INFO_LINK_MASK; depth = (pkg->link_info & ICM_LINK_INFO_DEPTH_MASK) >> ICM_LINK_INFO_DEPTH_SHIFT; @@ -1086,7 +1137,8 @@ static int icm_tr_disconnect_xdomain_paths(struct tb *tb, struct tb_xdomain *xd) } static void -icm_tr_device_connected(struct tb *tb, const struct icm_pkg_header *hdr) +__icm_tr_device_connected(struct tb *tb, const struct icm_pkg_header *hdr, + bool force_rtd3) { const struct icm_tr_event_device_connected *pkg = (const struct icm_tr_event_device_connected *)hdr; @@ -1096,6 +1148,8 @@ icm_tr_device_connected(struct tb *tb, const struct icm_pkg_header *hdr) bool authorized, boot; u64 route; + icm_postpone_rescan(tb); + /* * Currently we don't use the QoS information coming with the * device connected message so simply just ignore that extra @@ -1151,13 +1205,21 @@ icm_tr_device_connected(struct tb *tb, const struct icm_pkg_header *hdr) return; } - add_switch(parent_sw, route, &pkg->ep_uuid, (const u8 *)pkg->ep_name, - sizeof(pkg->ep_name), pkg->connection_id, - 0, 0, 0, security_level, authorized, boot); + sw = add_switch(parent_sw, route, &pkg->ep_uuid, (const u8 *)pkg->ep_name, + sizeof(pkg->ep_name), pkg->connection_id, 0, 0, 0, + security_level, authorized, boot); + if (!IS_ERR(sw) && force_rtd3) + sw->rpm = true; tb_switch_put(parent_sw); } +static void +icm_tr_device_connected(struct tb *tb, const struct icm_pkg_header *hdr) +{ + __icm_tr_device_connected(tb, hdr, false); +} + static void icm_tr_device_disconnected(struct tb *tb, const struct icm_pkg_header *hdr) { @@ -1468,6 +1530,61 @@ static int icm_ar_set_boot_acl(struct tb *tb, const uuid_t *uuids, return 0; } +static int +icm_icl_driver_ready(struct tb *tb, enum tb_security_level *security_level, + size_t *nboot_acl, bool *rpm) +{ + struct icm_tr_pkg_driver_ready_response reply; + struct icm_pkg_driver_ready request = { + .hdr.code = ICM_DRIVER_READY, + }; + int ret; + + memset(&reply, 0, sizeof(reply)); + ret = icm_request(tb, &request, sizeof(request), &reply, sizeof(reply), + 1, 20000); + if (ret) + return ret; + + /* Ice Lake always supports RTD3 */ + if (rpm) + *rpm = true; + + return 0; +} + +static void icm_icl_set_uuid(struct tb *tb) +{ + struct tb_nhi *nhi = tb->nhi; + u32 uuid[4]; + + pci_read_config_dword(nhi->pdev, VS_CAP_10, &uuid[0]); + pci_read_config_dword(nhi->pdev, VS_CAP_11, &uuid[1]); + uuid[2] = 0xffffffff; + uuid[3] = 0xffffffff; + + tb->root_switch->uuid = kmemdup(uuid, sizeof(uuid), GFP_KERNEL); +} + +static void +icm_icl_device_connected(struct tb *tb, const struct icm_pkg_header *hdr) +{ + __icm_tr_device_connected(tb, hdr, true); +} + +static void icm_icl_rtd3_veto(struct tb *tb, const struct icm_pkg_header *hdr) +{ + const struct icm_icl_event_rtd3_veto *pkg = + (const struct icm_icl_event_rtd3_veto *)hdr; + + tb_dbg(tb, "ICM rtd3 veto=0x%08x\n", pkg->veto_reason); + + if (pkg->veto_reason) + icm_veto_begin(tb); + else + icm_veto_end(tb); +} + static void icm_handle_notification(struct work_struct *work) { struct icm_notification *n = container_of(work, typeof(*n), work); @@ -1495,6 +1612,9 @@ static void icm_handle_notification(struct work_struct *work) case ICM_EVENT_XDOMAIN_DISCONNECTED: icm->xdomain_disconnected(tb, n->pkg); break; + case ICM_EVENT_RTD3_VETO: + icm->rtd3_veto(tb, n->pkg); + break; } } @@ -1853,6 +1973,13 @@ static void icm_complete(struct tb *tb) if (tb->nhi->going_away) return; + /* + * If RTD3 was vetoed before we entered system suspend allow it + * again now before driver ready is sent. Firmware sends a new RTD3 + * veto if it is still the case after we have sent it driver ready + * command. + */ + icm_veto_end(tb); icm_unplug_children(tb->root_switch); /* @@ -1918,6 +2045,9 @@ static int icm_start(struct tb *tb) tb->root_switch->no_nvm_upgrade = !icm->can_upgrade_nvm; tb->root_switch->rpm = icm->rpm; + if (icm->set_uuid) + icm->set_uuid(tb); + ret = tb_switch_add(tb->root_switch); if (ret) { tb_switch_put(tb->root_switch); @@ -2002,6 +2132,19 @@ static const struct tb_cm_ops icm_tr_ops = { .disconnect_xdomain_paths = icm_tr_disconnect_xdomain_paths, }; +/* Ice Lake */ +static const struct tb_cm_ops icm_icl_ops = { + .driver_ready = icm_driver_ready, + .start = icm_start, + .stop = icm_stop, + .complete = icm_complete, + .runtime_suspend = icm_runtime_suspend, + .runtime_resume = icm_runtime_resume, + .handle_event = icm_handle_event, + .approve_xdomain_paths = icm_tr_approve_xdomain_paths, + .disconnect_xdomain_paths = icm_tr_disconnect_xdomain_paths, +}; + struct tb *icm_probe(struct tb_nhi *nhi) { struct icm *icm; @@ -2070,6 +2213,19 @@ struct tb *icm_probe(struct tb_nhi *nhi) icm->xdomain_disconnected = icm_tr_xdomain_disconnected; tb->cm_ops = &icm_tr_ops; break; + + case PCI_DEVICE_ID_INTEL_ICL_NHI0: + case PCI_DEVICE_ID_INTEL_ICL_NHI1: + icm->is_supported = icm_ar_is_supported; + icm->driver_ready = icm_icl_driver_ready; + icm->set_uuid = icm_icl_set_uuid; + icm->device_connected = icm_icl_device_connected; + icm->device_disconnected = icm_tr_device_disconnected; + icm->xdomain_connected = icm_tr_xdomain_connected; + icm->xdomain_disconnected = icm_tr_xdomain_disconnected; + icm->rtd3_veto = icm_icl_rtd3_veto; + tb->cm_ops = &icm_icl_ops; + break; } if (!icm->is_supported || !icm->is_supported(tb)) { diff --git a/drivers/thunderbolt/nhi.c b/drivers/thunderbolt/nhi.c index 9c782706e652..641b21b54460 100644 --- a/drivers/thunderbolt/nhi.c +++ b/drivers/thunderbolt/nhi.c @@ -16,6 +16,7 @@ #include #include #include +#include #include "nhi.h" #include "nhi_regs.h" @@ -859,12 +860,52 @@ static irqreturn_t nhi_msi(int irq, void *data) return IRQ_HANDLED; } -static int nhi_suspend_noirq(struct device *dev) +static int __nhi_suspend_noirq(struct device *dev, bool wakeup) { struct pci_dev *pdev = to_pci_dev(dev); struct tb *tb = pci_get_drvdata(pdev); + struct tb_nhi *nhi = tb->nhi; + int ret; + + ret = tb_domain_suspend_noirq(tb); + if (ret) + return ret; + + if (nhi->ops && nhi->ops->suspend_noirq) { + ret = nhi->ops->suspend_noirq(tb->nhi, wakeup); + if (ret) + return ret; + } + + return 0; +} + +static int nhi_suspend_noirq(struct device *dev) +{ + return __nhi_suspend_noirq(dev, device_may_wakeup(dev)); +} + +static bool nhi_wake_supported(struct pci_dev *pdev) +{ + u8 val; + + /* + * If power rails are sustainable for wakeup from S4 this + * property is set by the BIOS. + */ + if (device_property_read_u8(&pdev->dev, "WAKE_SUPPORTED", &val)) + return !!val; + + return true; +} + +static int nhi_poweroff_noirq(struct device *dev) +{ + struct pci_dev *pdev = to_pci_dev(dev); + bool wakeup; - return tb_domain_suspend_noirq(tb); + wakeup = device_may_wakeup(dev) && nhi_wake_supported(pdev); + return __nhi_suspend_noirq(dev, wakeup); } static void nhi_enable_int_throttling(struct tb_nhi *nhi) @@ -887,16 +928,24 @@ static int nhi_resume_noirq(struct device *dev) { struct pci_dev *pdev = to_pci_dev(dev); struct tb *tb = pci_get_drvdata(pdev); + struct tb_nhi *nhi = tb->nhi; + int ret; /* * Check that the device is still there. It may be that the user * unplugged last device which causes the host controller to go * away on PCs. */ - if (!pci_device_is_present(pdev)) - tb->nhi->going_away = true; - else + if (!pci_device_is_present(pdev)) { + nhi->going_away = true; + } else { + if (nhi->ops && nhi->ops->resume_noirq) { + ret = nhi->ops->resume_noirq(nhi); + if (ret) + return ret; + } nhi_enable_int_throttling(tb->nhi); + } return tb_domain_resume_noirq(tb); } @@ -929,16 +978,35 @@ static int nhi_runtime_suspend(struct device *dev) { struct pci_dev *pdev = to_pci_dev(dev); struct tb *tb = pci_get_drvdata(pdev); + struct tb_nhi *nhi = tb->nhi; + int ret; + + ret = tb_domain_runtime_suspend(tb); + if (ret) + return ret; - return tb_domain_runtime_suspend(tb); + if (nhi->ops && nhi->ops->runtime_suspend) { + ret = nhi->ops->runtime_suspend(tb->nhi); + if (ret) + return ret; + } + return 0; } static int nhi_runtime_resume(struct device *dev) { struct pci_dev *pdev = to_pci_dev(dev); struct tb *tb = pci_get_drvdata(pdev); + struct tb_nhi *nhi = tb->nhi; + int ret; + + if (nhi->ops && nhi->ops->runtime_resume) { + ret = nhi->ops->runtime_resume(nhi); + if (ret) + return ret; + } - nhi_enable_int_throttling(tb->nhi); + nhi_enable_int_throttling(nhi); return tb_domain_runtime_resume(tb); } @@ -966,6 +1034,9 @@ static void nhi_shutdown(struct tb_nhi *nhi) flush_work(&nhi->interrupt_work); } ida_destroy(&nhi->msix_ida); + + if (nhi->ops && nhi->ops->shutdown) + nhi->ops->shutdown(nhi); } static int nhi_init_msi(struct tb_nhi *nhi) @@ -1010,12 +1081,27 @@ static int nhi_init_msi(struct tb_nhi *nhi) return 0; } +static bool nhi_imr_valid(struct pci_dev *pdev) +{ + u8 val; + + if (!device_property_read_u8(&pdev->dev, "IMR_VALID", &val)) + return !!val; + + return true; +} + static int nhi_probe(struct pci_dev *pdev, const struct pci_device_id *id) { struct tb_nhi *nhi; struct tb *tb; int res; + if (!nhi_imr_valid(pdev)) { + dev_warn(&pdev->dev, "firmware image not valid, aborting\n"); + return -ENODEV; + } + res = pcim_enable_device(pdev); if (res) { dev_err(&pdev->dev, "cannot enable PCI device, aborting\n"); @@ -1033,6 +1119,7 @@ static int nhi_probe(struct pci_dev *pdev, const struct pci_device_id *id) return -ENOMEM; nhi->pdev = pdev; + nhi->ops = (const struct tb_nhi_ops *)id->driver_data; /* cannot fail - table is allocated bin pcim_iomap_regions */ nhi->iobase = pcim_iomap_table(pdev)[0]; nhi->hop_count = ioread32(nhi->iobase + REG_HOP_COUNT) & 0x3ff; @@ -1065,6 +1152,12 @@ static int nhi_probe(struct pci_dev *pdev, const struct pci_device_id *id) pci_set_master(pdev); + if (nhi->ops && nhi->ops->init) { + res = nhi->ops->init(nhi); + if (res) + return res; + } + tb = icm_probe(nhi); if (!tb) tb = tb_probe(nhi); @@ -1125,6 +1218,7 @@ static const struct dev_pm_ops nhi_pm_ops = { .restore_noirq = nhi_resume_noirq, .suspend = nhi_suspend, .freeze = nhi_suspend, + .poweroff_noirq = nhi_poweroff_noirq, .poweroff = nhi_suspend, .complete = nhi_complete, .runtime_suspend = nhi_runtime_suspend, @@ -1172,6 +1266,10 @@ static struct pci_device_id nhi_ids[] = { { PCI_VDEVICE(INTEL, PCI_DEVICE_ID_INTEL_ALPINE_RIDGE_C_USBONLY_NHI) }, { PCI_VDEVICE(INTEL, PCI_DEVICE_ID_INTEL_TITAN_RIDGE_2C_NHI) }, { PCI_VDEVICE(INTEL, PCI_DEVICE_ID_INTEL_TITAN_RIDGE_4C_NHI) }, + { PCI_VDEVICE(INTEL, PCI_DEVICE_ID_INTEL_ICL_NHI0), + .driver_data = (kernel_ulong_t)&icl_nhi_ops }, + { PCI_VDEVICE(INTEL, PCI_DEVICE_ID_INTEL_ICL_NHI1), + .driver_data = (kernel_ulong_t)&icl_nhi_ops }, { 0,} }; diff --git a/drivers/thunderbolt/nhi.h b/drivers/thunderbolt/nhi.h index 1b5d47ecd3ed..b7b973949f8e 100644 --- a/drivers/thunderbolt/nhi.h +++ b/drivers/thunderbolt/nhi.h @@ -30,6 +30,26 @@ enum nhi_mailbox_cmd { int nhi_mailbox_cmd(struct tb_nhi *nhi, enum nhi_mailbox_cmd cmd, u32 data); enum nhi_fw_mode nhi_mailbox_mode(struct tb_nhi *nhi); +/** + * struct tb_nhi_ops - NHI specific optional operations + * @init: NHI specific initialization + * @suspend_noirq: NHI specific suspend_noirq hook + * @resume_noirq: NHI specific resume_noirq hook + * @runtime_suspend: NHI specific runtime_suspend hook + * @runtime_resume: NHI specific runtime_resume hook + * @shutdown: NHI specific shutdown + */ +struct tb_nhi_ops { + int (*init)(struct tb_nhi *nhi); + int (*suspend_noirq)(struct tb_nhi *nhi, bool wakeup); + int (*resume_noirq)(struct tb_nhi *nhi); + int (*runtime_suspend)(struct tb_nhi *nhi); + int (*runtime_resume)(struct tb_nhi *nhi); + void (*shutdown)(struct tb_nhi *nhi); +}; + +extern const struct tb_nhi_ops icl_nhi_ops; + /* * PCI IDs used in this driver from Win Ridge forward. There is no * need for the PCI quirk anymore as we will use ICM also on Apple @@ -51,5 +71,7 @@ enum nhi_fw_mode nhi_mailbox_mode(struct tb_nhi *nhi); #define PCI_DEVICE_ID_INTEL_TITAN_RIDGE_4C_BRIDGE 0x15ea #define PCI_DEVICE_ID_INTEL_TITAN_RIDGE_4C_NHI 0x15eb #define PCI_DEVICE_ID_INTEL_TITAN_RIDGE_DD_BRIDGE 0x15ef +#define PCI_DEVICE_ID_INTEL_ICL_NHI1 0x8a0d +#define PCI_DEVICE_ID_INTEL_ICL_NHI0 0x8a17 #endif diff --git a/drivers/thunderbolt/nhi_ops.c b/drivers/thunderbolt/nhi_ops.c new file mode 100644 index 000000000000..61cd09cef943 --- /dev/null +++ b/drivers/thunderbolt/nhi_ops.c @@ -0,0 +1,179 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * NHI specific operations + * + * Copyright (C) 2019, Intel Corporation + * Author: Mika Westerberg + */ + +#include +#include + +#include "nhi.h" +#include "nhi_regs.h" +#include "tb.h" + +/* Ice Lake specific NHI operations */ + +#define ICL_LC_MAILBOX_TIMEOUT 500 /* ms */ + +static int check_for_device(struct device *dev, void *data) +{ + return tb_is_switch(dev); +} + +static bool icl_nhi_is_device_connected(struct tb_nhi *nhi) +{ + struct tb *tb = pci_get_drvdata(nhi->pdev); + int ret; + + ret = device_for_each_child(&tb->root_switch->dev, NULL, + check_for_device); + return ret > 0; +} + +static int icl_nhi_force_power(struct tb_nhi *nhi, bool power) +{ + u32 vs_cap; + + /* + * The Thunderbolt host controller is present always in Ice Lake + * but the firmware may not be loaded and running (depending + * whether there is device connected and so on). Each time the + * controller is used we need to "Force Power" it first and wait + * for the firmware to indicate it is up and running. This "Force + * Power" is really not about actually powering on/off the + * controller so it is accessible even if "Force Power" is off. + * + * The actual power management happens inside shared ACPI power + * resources using standard ACPI methods. + */ + pci_read_config_dword(nhi->pdev, VS_CAP_22, &vs_cap); + if (power) { + vs_cap &= ~VS_CAP_22_DMA_DELAY_MASK; + vs_cap |= 0x22 << VS_CAP_22_DMA_DELAY_SHIFT; + vs_cap |= VS_CAP_22_FORCE_POWER; + } else { + vs_cap &= ~VS_CAP_22_FORCE_POWER; + } + pci_write_config_dword(nhi->pdev, VS_CAP_22, vs_cap); + + if (power) { + unsigned int retries = 10; + u32 val; + + /* Wait until the firmware tells it is up and running */ + do { + pci_read_config_dword(nhi->pdev, VS_CAP_9, &val); + if (val & VS_CAP_9_FW_READY) + return 0; + msleep(250); + } while (--retries); + + return -ETIMEDOUT; + } + + return 0; +} + +static void icl_nhi_lc_mailbox_cmd(struct tb_nhi *nhi, enum icl_lc_mailbox_cmd cmd) +{ + u32 data; + + pci_read_config_dword(nhi->pdev, VS_CAP_19, &data); + data = (cmd << VS_CAP_19_CMD_SHIFT) & VS_CAP_19_CMD_MASK; + pci_write_config_dword(nhi->pdev, VS_CAP_19, data | VS_CAP_19_VALID); +} + +static int icl_nhi_lc_mailbox_cmd_complete(struct tb_nhi *nhi, int timeout) +{ + unsigned long end; + u32 data; + + if (!timeout) + goto clear; + + end = jiffies + msecs_to_jiffies(timeout); + do { + pci_read_config_dword(nhi->pdev, VS_CAP_18, &data); + if (data & VS_CAP_18_DONE) + goto clear; + msleep(100); + } while (time_before(jiffies, end)); + + return -ETIMEDOUT; + +clear: + /* Clear the valid bit */ + pci_write_config_dword(nhi->pdev, VS_CAP_19, 0); + return 0; +} + +static void icl_nhi_set_ltr(struct tb_nhi *nhi) +{ + u32 max_ltr, ltr; + + pci_read_config_dword(nhi->pdev, VS_CAP_16, &max_ltr); + max_ltr &= 0xffff; + /* Program the same value for both snoop and no-snoop */ + ltr = max_ltr << 16 | max_ltr; + pci_write_config_dword(nhi->pdev, VS_CAP_15, ltr); +} + +static int icl_nhi_suspend(struct tb_nhi *nhi) +{ + int ret; + + if (icl_nhi_is_device_connected(nhi)) + return 0; + + /* + * If there is no device connected we need to perform both: a + * handshake through LC mailbox and force power down before + * entering D3. + */ + icl_nhi_lc_mailbox_cmd(nhi, ICL_LC_PREPARE_FOR_RESET); + ret = icl_nhi_lc_mailbox_cmd_complete(nhi, ICL_LC_MAILBOX_TIMEOUT); + if (ret) + return ret; + + return icl_nhi_force_power(nhi, false); +} + +static int icl_nhi_suspend_noirq(struct tb_nhi *nhi, bool wakeup) +{ + enum icl_lc_mailbox_cmd cmd; + + if (!pm_suspend_via_firmware()) + return icl_nhi_suspend(nhi); + + cmd = wakeup ? ICL_LC_GO2SX : ICL_LC_GO2SX_NO_WAKE; + icl_nhi_lc_mailbox_cmd(nhi, cmd); + return icl_nhi_lc_mailbox_cmd_complete(nhi, ICL_LC_MAILBOX_TIMEOUT); +} + +static int icl_nhi_resume(struct tb_nhi *nhi) +{ + int ret; + + ret = icl_nhi_force_power(nhi, true); + if (ret) + return ret; + + icl_nhi_set_ltr(nhi); + return 0; +} + +static void icl_nhi_shutdown(struct tb_nhi *nhi) +{ + icl_nhi_force_power(nhi, false); +} + +const struct tb_nhi_ops icl_nhi_ops = { + .init = icl_nhi_resume, + .suspend_noirq = icl_nhi_suspend_noirq, + .resume_noirq = icl_nhi_resume, + .runtime_suspend = icl_nhi_suspend, + .runtime_resume = icl_nhi_resume, + .shutdown = icl_nhi_shutdown, +}; diff --git a/drivers/thunderbolt/nhi_regs.h b/drivers/thunderbolt/nhi_regs.h index a60bd98c1d04..0d4970dcef84 100644 --- a/drivers/thunderbolt/nhi_regs.h +++ b/drivers/thunderbolt/nhi_regs.h @@ -124,4 +124,41 @@ struct ring_desc { #define REG_FW_STS_ICM_EN_INVERT BIT(1) #define REG_FW_STS_ICM_EN BIT(0) +/* ICL NHI VSEC registers */ + +/* FW ready */ +#define VS_CAP_9 0xc8 +#define VS_CAP_9_FW_READY BIT(31) +/* UUID */ +#define VS_CAP_10 0xcc +#define VS_CAP_11 0xd0 +/* LTR */ +#define VS_CAP_15 0xe0 +#define VS_CAP_16 0xe4 +/* TBT2PCIe */ +#define VS_CAP_18 0xec +#define VS_CAP_18_DONE BIT(0) +/* PCIe2TBT */ +#define VS_CAP_19 0xf0 +#define VS_CAP_19_VALID BIT(0) +#define VS_CAP_19_CMD_SHIFT 1 +#define VS_CAP_19_CMD_MASK GENMASK(7, 1) +/* Force power */ +#define VS_CAP_22 0xfc +#define VS_CAP_22_FORCE_POWER BIT(1) +#define VS_CAP_22_DMA_DELAY_MASK GENMASK(31, 24) +#define VS_CAP_22_DMA_DELAY_SHIFT 24 + +/** + * enum icl_lc_mailbox_cmd - ICL specific LC mailbox commands + * @ICL_LC_GO2SX: Ask LC to enter Sx without wake + * @ICL_LC_GO2SX_NO_WAKE: Ask LC to enter Sx with wake + * @ICL_LC_PREPARE_FOR_RESET: Prepare LC for reset + */ +enum icl_lc_mailbox_cmd { + ICL_LC_GO2SX = 0x02, + ICL_LC_GO2SX_NO_WAKE = 0x03, + ICL_LC_PREPARE_FOR_RESET = 0x21, +}; + #endif diff --git a/drivers/thunderbolt/switch.c b/drivers/thunderbolt/switch.c index 531f11fecf75..410bf1bceeee 100644 --- a/drivers/thunderbolt/switch.c +++ b/drivers/thunderbolt/switch.c @@ -1470,6 +1470,8 @@ static int tb_switch_get_generation(struct tb_switch *sw) case PCI_DEVICE_ID_INTEL_TITAN_RIDGE_2C_BRIDGE: case PCI_DEVICE_ID_INTEL_TITAN_RIDGE_4C_BRIDGE: case PCI_DEVICE_ID_INTEL_TITAN_RIDGE_DD_BRIDGE: + case PCI_DEVICE_ID_INTEL_ICL_NHI0: + case PCI_DEVICE_ID_INTEL_ICL_NHI1: return 3; default: diff --git a/drivers/thunderbolt/tb_msgs.h b/drivers/thunderbolt/tb_msgs.h index afbe1d29bb03..4b641e4ee0c5 100644 --- a/drivers/thunderbolt/tb_msgs.h +++ b/drivers/thunderbolt/tb_msgs.h @@ -104,10 +104,11 @@ enum icm_pkg_code { }; enum icm_event_code { - ICM_EVENT_DEVICE_CONNECTED = 3, - ICM_EVENT_DEVICE_DISCONNECTED = 4, - ICM_EVENT_XDOMAIN_CONNECTED = 6, - ICM_EVENT_XDOMAIN_DISCONNECTED = 7, + ICM_EVENT_DEVICE_CONNECTED = 0x3, + ICM_EVENT_DEVICE_DISCONNECTED = 0x4, + ICM_EVENT_XDOMAIN_CONNECTED = 0x6, + ICM_EVENT_XDOMAIN_DISCONNECTED = 0x7, + ICM_EVENT_RTD3_VETO = 0xa, }; struct icm_pkg_header { @@ -463,6 +464,13 @@ struct icm_tr_pkg_disconnect_xdomain_response { uuid_t remote_uuid; }; +/* Ice Lake messages */ + +struct icm_icl_event_rtd3_veto { + struct icm_pkg_header hdr; + u32 veto_reason; +}; + /* XDomain messages */ struct tb_xdomain_header { diff --git a/include/linux/thunderbolt.h b/include/linux/thunderbolt.h index 2d7e012db03f..ece782ef5466 100644 --- a/include/linux/thunderbolt.h +++ b/include/linux/thunderbolt.h @@ -429,6 +429,7 @@ static inline struct tb_xdomain *tb_service_parent(struct tb_service *svc) * @lock: Must be held during ring creation/destruction. Is acquired by * interrupt_work when dispatching interrupts to individual rings. * @pdev: Pointer to the PCI device + * @ops: NHI specific optional ops * @iobase: MMIO space of the NHI * @tx_rings: All Tx rings available on this host controller * @rx_rings: All Rx rings available on this host controller @@ -442,6 +443,7 @@ static inline struct tb_xdomain *tb_service_parent(struct tb_service *svc) struct tb_nhi { spinlock_t lock; struct pci_dev *pdev; + const struct tb_nhi_ops *ops; void __iomem *iobase; struct tb_ring **tx_rings; struct tb_ring **rx_rings; -- cgit v1.2.3-59-g8ed1b From b5dc75c915cdaebab9b9875022e45638d6b14a7e Mon Sep 17 00:00:00 2001 From: Richard Gong Date: Tue, 3 Sep 2019 08:18:18 -0500 Subject: firmware: stratix10-svc: extend svc to support new RSU features Extend Intel Stratix10 service layer driver to support new RSU notify and MAX_RETRY with watchdog event. RSU is used to provide our customers with protection against loading bad bitstream onto their devices when those devices are booting from flash RSU notifies provides users with an API to notify the firmware of the state of hard processor system. To deal with watchdog event, RSU provides a way for user to retry the current running image several times before giving up and starting normal RSU failover flow. Signed-off-by: Richard Gong Reviewed-by: Alan Tull Link: https://lore.kernel.org/r/1567516701-26026-2-git-send-email-richard.gong@linux.intel.com Signed-off-by: Greg Kroah-Hartman --- drivers/firmware/stratix10-svc.c | 76 +++++++++++++++++++++- include/linux/firmware/intel/stratix10-smc.h | 51 +++++++++++++-- .../linux/firmware/intel/stratix10-svc-client.h | 11 +++- 3 files changed, 128 insertions(+), 10 deletions(-) (limited to 'include') diff --git a/drivers/firmware/stratix10-svc.c b/drivers/firmware/stratix10-svc.c index 6e6514825ad0..b485321189e1 100644 --- a/drivers/firmware/stratix10-svc.c +++ b/drivers/firmware/stratix10-svc.c @@ -38,12 +38,23 @@ #define FPGA_CONFIG_DATA_CLAIM_TIMEOUT_MS 200 #define FPGA_CONFIG_STATUS_TIMEOUT_SEC 30 +/* stratix10 service layer clients */ +#define STRATIX10_RSU "stratix10-rsu" + typedef void (svc_invoke_fn)(unsigned long, unsigned long, unsigned long, unsigned long, unsigned long, unsigned long, unsigned long, unsigned long, struct arm_smccc_res *); struct stratix10_svc_chan; +/** + * struct stratix10_svc - svc private data + * @stratix10_svc_rsu: pointer to stratix10 RSU device + */ +struct stratix10_svc { + struct platform_device *stratix10_svc_rsu; +}; + /** * struct stratix10_svc_sh_memory - service shared memory structure * @sync_complete: state for a completion @@ -296,7 +307,12 @@ static void svc_thread_recv_status_ok(struct stratix10_svc_data *p_data, cb_data->status = BIT(SVC_STATUS_RECONFIG_COMPLETED); break; case COMMAND_RSU_UPDATE: + case COMMAND_RSU_NOTIFY: + cb_data->status = BIT(SVC_STATUS_RSU_OK); + break; + case COMMAND_RSU_RETRY: cb_data->status = BIT(SVC_STATUS_RSU_OK); + cb_data->kaddr1 = &res.a1; break; default: pr_warn("it shouldn't happen\n"); @@ -386,6 +402,16 @@ static int svc_normal_to_secure_thread(void *data) a1 = pdata->arg[0]; a2 = 0; break; + case COMMAND_RSU_NOTIFY: + a0 = INTEL_SIP_SMC_RSU_NOTIFY; + a1 = pdata->arg[0]; + a2 = 0; + break; + case COMMAND_RSU_RETRY: + a0 = INTEL_SIP_SMC_RSU_RETRY_COUNTER; + a1 = 0; + a2 = 0; + break; default: pr_warn("it shouldn't happen\n"); break; @@ -438,7 +464,28 @@ static int svc_normal_to_secure_thread(void *data) pr_debug("%s: STATUS_REJECTED\n", __func__); break; case INTEL_SIP_SMC_FPGA_CONFIG_STATUS_ERROR: + case INTEL_SIP_SMC_RSU_ERROR: pr_err("%s: STATUS_ERROR\n", __func__); + switch (pdata->command) { + /* for FPGA mgr */ + case COMMAND_RECONFIG_DATA_CLAIM: + case COMMAND_RECONFIG: + case COMMAND_RECONFIG_DATA_SUBMIT: + case COMMAND_RECONFIG_STATUS: + cbdata->status = + BIT(SVC_STATUS_RECONFIG_ERROR); + break; + + /* for RSU */ + case COMMAND_RSU_STATUS: + case COMMAND_RSU_UPDATE: + case COMMAND_RSU_NOTIFY: + case COMMAND_RSU_RETRY: + cbdata->status = + BIT(SVC_STATUS_RSU_ERROR); + break; + } + cbdata->status = BIT(SVC_STATUS_RECONFIG_ERROR); cbdata->kaddr1 = NULL; cbdata->kaddr2 = NULL; @@ -530,7 +577,7 @@ static int svc_get_sh_memory(struct platform_device *pdev, if (!sh_memory->addr || !sh_memory->size) { dev_err(dev, - "fails to get shared memory info from secure world\n"); + "failed to get shared memory info from secure world\n"); return -ENOMEM; } @@ -768,7 +815,7 @@ int stratix10_svc_send(struct stratix10_svc_chan *chan, void *msg) "svc_smc_hvc_thread"); if (IS_ERR(chan->ctrl->task)) { dev_err(chan->ctrl->dev, - "fails to create svc_smc_hvc_thread\n"); + "failed to create svc_smc_hvc_thread\n"); kfree(p_data); return -EINVAL; } @@ -913,6 +960,8 @@ static int stratix10_svc_drv_probe(struct platform_device *pdev) struct stratix10_svc_chan *chans; struct gen_pool *genpool; struct stratix10_svc_sh_memory *sh_memory; + struct stratix10_svc *svc; + svc_invoke_fn *invoke_fn; size_t fifo_size; int ret; @@ -957,7 +1006,7 @@ static int stratix10_svc_drv_probe(struct platform_device *pdev) fifo_size = sizeof(struct stratix10_svc_data) * SVC_NUM_DATA_IN_FIFO; ret = kfifo_alloc(&controller->svc_fifo, fifo_size, GFP_KERNEL); if (ret) { - dev_err(dev, "fails to allocate FIFO\n"); + dev_err(dev, "failed to allocate FIFO\n"); return ret; } spin_lock_init(&controller->svc_fifo_lock); @@ -975,6 +1024,24 @@ static int stratix10_svc_drv_probe(struct platform_device *pdev) list_add_tail(&controller->node, &svc_ctrl); platform_set_drvdata(pdev, controller); + /* add svc client device(s) */ + svc = devm_kzalloc(dev, sizeof(*svc), GFP_KERNEL); + if (!svc) + return -ENOMEM; + + svc->stratix10_svc_rsu = platform_device_alloc(STRATIX10_RSU, 0); + if (!svc->stratix10_svc_rsu) { + dev_err(dev, "failed to allocate %s device\n", STRATIX10_RSU); + return -ENOMEM; + } + + ret = platform_device_add(svc->stratix10_svc_rsu); + if (ret) { + platform_device_put(svc->stratix10_svc_rsu); + return ret; + } + dev_set_drvdata(dev, svc); + pr_info("Intel Service Layer Driver Initialized\n"); return ret; @@ -982,8 +1049,11 @@ static int stratix10_svc_drv_probe(struct platform_device *pdev) static int stratix10_svc_drv_remove(struct platform_device *pdev) { + struct stratix10_svc *svc = dev_get_drvdata(&pdev->dev); struct stratix10_svc_controller *ctrl = platform_get_drvdata(pdev); + platform_device_unregister(svc->stratix10_svc_rsu); + kfifo_free(&ctrl->svc_fifo); if (ctrl->task) { kthread_stop(ctrl->task); diff --git a/include/linux/firmware/intel/stratix10-smc.h b/include/linux/firmware/intel/stratix10-smc.h index 01684d935580..013ae4819deb 100644 --- a/include/linux/firmware/intel/stratix10-smc.h +++ b/include/linux/firmware/intel/stratix10-smc.h @@ -210,7 +210,7 @@ INTEL_SIP_SMC_FAST_CALL_VAL(INTEL_SIP_SMC_FUNCID_FPGA_CONFIG_COMPLETED_WRITE) #define INTEL_SIP_SMC_FPGA_CONFIG_LOOPBACK \ INTEL_SIP_SMC_FAST_CALL_VAL(INTEL_SIP_SMC_FUNCID_FPGA_CONFIG_LOOPBACK) -/* +/** * Request INTEL_SIP_SMC_REG_READ * * Read a protected register at EL3 @@ -229,7 +229,7 @@ INTEL_SIP_SMC_FAST_CALL_VAL(INTEL_SIP_SMC_FUNCID_FPGA_CONFIG_COMPLETED_WRITE) #define INTEL_SIP_SMC_REG_READ \ INTEL_SIP_SMC_FAST_CALL_VAL(INTEL_SIP_SMC_FUNCID_REG_READ) -/* +/** * Request INTEL_SIP_SMC_REG_WRITE * * Write a protected register at EL3 @@ -248,7 +248,7 @@ INTEL_SIP_SMC_FAST_CALL_VAL(INTEL_SIP_SMC_FUNCID_FPGA_CONFIG_COMPLETED_WRITE) #define INTEL_SIP_SMC_REG_WRITE \ INTEL_SIP_SMC_FAST_CALL_VAL(INTEL_SIP_SMC_FUNCID_REG_WRITE) -/* +/** * Request INTEL_SIP_SMC_FUNCID_REG_UPDATE * * Update one or more bits in a protected register at EL3 using a @@ -269,7 +269,7 @@ INTEL_SIP_SMC_FAST_CALL_VAL(INTEL_SIP_SMC_FUNCID_FPGA_CONFIG_COMPLETED_WRITE) #define INTEL_SIP_SMC_REG_UPDATE \ INTEL_SIP_SMC_FAST_CALL_VAL(INTEL_SIP_SMC_FUNCID_REG_UPDATE) -/* +/** * Request INTEL_SIP_SMC_RSU_STATUS * * Request remote status update boot log, call is synchronous. @@ -292,7 +292,7 @@ INTEL_SIP_SMC_FAST_CALL_VAL(INTEL_SIP_SMC_FUNCID_FPGA_CONFIG_COMPLETED_WRITE) #define INTEL_SIP_SMC_RSU_STATUS \ INTEL_SIP_SMC_FAST_CALL_VAL(INTEL_SIP_SMC_FUNCID_RSU_STATUS) -/* +/** * Request INTEL_SIP_SMC_RSU_UPDATE * * Request to set the offset of the bitstream to boot after reboot, call @@ -310,7 +310,7 @@ INTEL_SIP_SMC_FAST_CALL_VAL(INTEL_SIP_SMC_FUNCID_FPGA_CONFIG_COMPLETED_WRITE) #define INTEL_SIP_SMC_RSU_UPDATE \ INTEL_SIP_SMC_FAST_CALL_VAL(INTEL_SIP_SMC_FUNCID_RSU_UPDATE) -/* +/** * Request INTEL_SIP_SMC_ECC_DBE * * Sync call used by service driver at EL1 to alert EL3 that a Double @@ -329,3 +329,42 @@ INTEL_SIP_SMC_FAST_CALL_VAL(INTEL_SIP_SMC_FUNCID_FPGA_CONFIG_COMPLETED_WRITE) INTEL_SIP_SMC_FAST_CALL_VAL(INTEL_SIP_SMC_FUNCID_ECC_DBE) #endif + +/** + * Request INTEL_SIP_SMC_RSU_NOTIFY + * + * Sync call used by service driver at EL1 to report hard processor + * system execution stage to firmware + * + * Call register usage: + * a0 INTEL_SIP_SMC_RSU_NOTIFY + * a1 32bit value representing hard processor system execution stage + * a2-7 not used + * + * Return status + * a0 INTEL_SIP_SMC_STATUS_OK + */ +#define INTEL_SIP_SMC_FUNCID_RSU_NOTIFY 14 +#define INTEL_SIP_SMC_RSU_NOTIFY \ + INTEL_SIP_SMC_FAST_CALL_VAL(INTEL_SIP_SMC_FUNCID_RSU_NOTIFY) + +/** + * Request INTEL_SIP_SMC_RSU_RETRY_COUNTER + * + * Sync call used by service driver at EL1 to query RSU retry counter + * + * Call register usage: + * a0 INTEL_SIP_SMC_RSU_RETRY_COUNTER + * a1-7 not used + * + * Return status + * a0 INTEL_SIP_SMC_STATUS_OK + * a1 the retry counter + * + * Or + * + * a0 INTEL_SIP_SMC_RSU_ERROR + */ +#define INTEL_SIP_SMC_FUNCID_RSU_RETRY_COUNTER 15 +#define INTEL_SIP_SMC_RSU_RETRY_COUNTER \ + INTEL_SIP_SMC_FAST_CALL_VAL(INTEL_SIP_SMC_FUNCID_RSU_RETRY_COUNTER) diff --git a/include/linux/firmware/intel/stratix10-svc-client.h b/include/linux/firmware/intel/stratix10-svc-client.h index e521f172a47a..b6c4302a39e0 100644 --- a/include/linux/firmware/intel/stratix10-svc-client.h +++ b/include/linux/firmware/intel/stratix10-svc-client.h @@ -95,6 +95,13 @@ struct stratix10_svc_chan; * * @COMMAND_RSU_UPDATE: set the offset of the bitstream to boot after reboot, * return status is SVC_STATUS_RSU_OK or SVC_STATUS_RSU_ERROR + * + * @COMMAND_RSU_NOTIFY: report the status of hard processor system + * software to firmware, return status is SVC_STATUS_RSU_OK or + * SVC_STATUS_RSU_ERROR + * + * @COMMAND_RSU_RETRY: query firmware for the current image's retry counter, + * return status is SVC_STATUS_RSU_OK or SVC_STATUS_RSU_ERROR */ enum stratix10_svc_command_code { COMMAND_NOOP = 0, @@ -103,7 +110,9 @@ enum stratix10_svc_command_code { COMMAND_RECONFIG_DATA_CLAIM, COMMAND_RECONFIG_STATUS, COMMAND_RSU_STATUS, - COMMAND_RSU_UPDATE + COMMAND_RSU_UPDATE, + COMMAND_RSU_NOTIFY, + COMMAND_RSU_RETRY, }; /** -- cgit v1.2.3-59-g8ed1b From 74ded38a8e185b57a4362787815c353ac595013c Mon Sep 17 00:00:00 2001 From: Thomas Bogendoerfer Date: Sat, 31 Aug 2019 10:26:21 +0200 Subject: w1: add 1-wire master driver for IP block found in SGI ASICs Starting with SGI Origin machines nearly every new SGI ASIC contains an 1-Wire master. They are used for attaching One-Wire prom devices, which contain information about part numbers, revision numbers, serial number etc. and MAC addresses for ethernet interfaces. This patch adds a master driver to support this IP block. It also adds an extra field dev_id to struct w1_bus_master, which could be in used in slave drivers for creating unique device names. Signed-off-by: Thomas Bogendoerfer Link: https://lore.kernel.org/r/20190831082623.15627-2-tbogendoerfer@suse.de Signed-off-by: Greg Kroah-Hartman --- drivers/w1/masters/Kconfig | 9 +++ drivers/w1/masters/Makefile | 1 + drivers/w1/masters/sgi_w1.c | 130 +++++++++++++++++++++++++++++++++++ include/linux/platform_data/sgi-w1.h | 13 ++++ include/linux/w1.h | 5 ++ 5 files changed, 158 insertions(+) create mode 100644 drivers/w1/masters/sgi_w1.c create mode 100644 include/linux/platform_data/sgi-w1.h (limited to 'include') diff --git a/drivers/w1/masters/Kconfig b/drivers/w1/masters/Kconfig index 7ae260577901..24b9a8e05f64 100644 --- a/drivers/w1/masters/Kconfig +++ b/drivers/w1/masters/Kconfig @@ -65,5 +65,14 @@ config HDQ_MASTER_OMAP Say Y here if you want support for the 1-wire or HDQ Interface on an OMAP processor. +config W1_MASTER_SGI + tristate "SGI ASIC driver" + help + Say Y here if you want support for your 1-wire devices using + SGI ASIC 1-Wire interface + + This support is also available as a module. If so, the module + will be called sgi_w1. + endmenu diff --git a/drivers/w1/masters/Makefile b/drivers/w1/masters/Makefile index 18954cae4256..dae629b7ab49 100644 --- a/drivers/w1/masters/Makefile +++ b/drivers/w1/masters/Makefile @@ -11,3 +11,4 @@ obj-$(CONFIG_W1_MASTER_MXC) += mxc_w1.o obj-$(CONFIG_W1_MASTER_DS1WM) += ds1wm.o obj-$(CONFIG_W1_MASTER_GPIO) += w1-gpio.o obj-$(CONFIG_HDQ_MASTER_OMAP) += omap_hdq.o +obj-$(CONFIG_W1_MASTER_SGI) += sgi_w1.o diff --git a/drivers/w1/masters/sgi_w1.c b/drivers/w1/masters/sgi_w1.c new file mode 100644 index 000000000000..1b2d96b945be --- /dev/null +++ b/drivers/w1/masters/sgi_w1.c @@ -0,0 +1,130 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * sgi_w1.c - w1 master driver for one wire support in SGI ASICs + */ + +#include +#include +#include +#include +#include +#include +#include +#include + +#include + +#define MCR_RD_DATA BIT(0) +#define MCR_DONE BIT(1) + +#define MCR_PACK(pulse, sample) (((pulse) << 10) | ((sample) << 2)) + +struct sgi_w1_device { + u32 __iomem *mcr; + struct w1_bus_master bus_master; + char dev_id[64]; +}; + +static u8 sgi_w1_wait(u32 __iomem *mcr) +{ + u32 mcr_val; + + do { + mcr_val = readl(mcr); + } while (!(mcr_val & MCR_DONE)); + + return (mcr_val & MCR_RD_DATA) ? 1 : 0; +} + +/* + * this is the low level routine to + * reset the device on the One Wire interface + * on the hardware + */ +static u8 sgi_w1_reset_bus(void *data) +{ + struct sgi_w1_device *dev = data; + u8 ret; + + writel(MCR_PACK(520, 65), dev->mcr); + ret = sgi_w1_wait(dev->mcr); + udelay(500); /* recovery time */ + return ret; +} + +/* + * this is the low level routine to read/write a bit on the One Wire + * interface on the hardware. It does write 0 if parameter bit is set + * to 0, otherwise a write 1/read. + */ +static u8 sgi_w1_touch_bit(void *data, u8 bit) +{ + struct sgi_w1_device *dev = data; + u8 ret; + + if (bit) + writel(MCR_PACK(6, 13), dev->mcr); + else + writel(MCR_PACK(80, 30), dev->mcr); + + ret = sgi_w1_wait(dev->mcr); + if (bit) + udelay(100); /* recovery */ + return ret; +} + +static int sgi_w1_probe(struct platform_device *pdev) +{ + struct sgi_w1_device *sdev; + struct sgi_w1_platform_data *pdata; + struct resource *res; + + sdev = devm_kzalloc(&pdev->dev, sizeof(struct sgi_w1_device), + GFP_KERNEL); + if (!sdev) + return -ENOMEM; + + res = platform_get_resource(pdev, IORESOURCE_MEM, 0); + sdev->mcr = devm_ioremap_resource(&pdev->dev, res); + if (IS_ERR(sdev->mcr)) + return PTR_ERR(sdev->mcr); + + sdev->bus_master.data = sdev; + sdev->bus_master.reset_bus = sgi_w1_reset_bus; + sdev->bus_master.touch_bit = sgi_w1_touch_bit; + + pdata = dev_get_platdata(&pdev->dev); + if (pdata) { + strlcpy(sdev->dev_id, pdata->dev_id, sizeof(sdev->dev_id)); + sdev->bus_master.dev_id = sdev->dev_id; + } + + platform_set_drvdata(pdev, sdev); + + return w1_add_master_device(&sdev->bus_master); +} + +/* + * disassociate the w1 device from the driver + */ +static int sgi_w1_remove(struct platform_device *pdev) +{ + struct sgi_w1_device *sdev = platform_get_drvdata(pdev); + + w1_remove_master_device(&sdev->bus_master); + + return 0; +} + +static struct platform_driver sgi_w1_driver = { + .driver = { + .name = "sgi_w1", + }, + .probe = sgi_w1_probe, + .remove = sgi_w1_remove, +}; +module_platform_driver(sgi_w1_driver); + +MODULE_LICENSE("GPL"); +MODULE_AUTHOR("Thomas Bogendoerfer"); +MODULE_DESCRIPTION("Driver for One-Wire IP in SGI ASICs"); diff --git a/include/linux/platform_data/sgi-w1.h b/include/linux/platform_data/sgi-w1.h new file mode 100644 index 000000000000..e28c8a90ff84 --- /dev/null +++ b/include/linux/platform_data/sgi-w1.h @@ -0,0 +1,13 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * SGI One-Wire (W1) IP + */ + +#ifndef PLATFORM_DATA_SGI_W1_H +#define PLATFORM_DATA_SGI_W1_H + +struct sgi_w1_platform_data { + char dev_id[64]; +}; + +#endif /* PLATFORM_DATA_SGI_W1_H */ diff --git a/include/linux/w1.h b/include/linux/w1.h index e0b5156f78fd..7da0c7588e04 100644 --- a/include/linux/w1.h +++ b/include/linux/w1.h @@ -118,6 +118,9 @@ typedef void (*w1_slave_found_callback)(struct w1_master *, u64); * w1_master* is passed to the slave found callback. * u8 is search_type, W1_SEARCH or W1_ALARM_SEARCH * + * @dev_id: Optional device id string, which w1 slaves could use for + * creating names, which then give a connection to the w1 master + * * Note: read_bit and write_bit are very low level functions and should only * be used with hardware that doesn't really support 1-wire operations, * like a parallel/serial port. @@ -150,6 +153,8 @@ struct w1_bus_master { void (*search)(void *, struct w1_master *, u8, w1_slave_found_callback); + + char *dev_id; }; /** -- cgit v1.2.3-59-g8ed1b From f388ec7c16ad8676ee516a735a6b7588252f971d Mon Sep 17 00:00:00 2001 From: Oded Gabbay Date: Tue, 16 Jul 2019 08:55:04 +0300 Subject: habanalabs: add comments on INFO IOCTL This patch adds some in-code documentation on the different opcodes of the INFO IOCTL. Signed-off-by: Oded Gabbay Reviewed-by: Omer Shpigelman --- include/uapi/misc/habanalabs.h | 14 +++++++++++++- 1 file changed, 13 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/uapi/misc/habanalabs.h b/include/uapi/misc/habanalabs.h index 3956c226ca35..a5a1d0e7ec82 100644 --- a/include/uapi/misc/habanalabs.h +++ b/include/uapi/misc/habanalabs.h @@ -75,7 +75,19 @@ enum hl_device_status { HL_DEVICE_STATUS_MALFUNCTION }; -/* Opcode for management ioctl */ +/* Opcode for management ioctl + * + * HW_IP_INFO - Receive information about different IP blocks in the + * device. + * HL_INFO_HW_EVENTS - Receive an array describing how many times each event + * occurred since the last hard reset. + * HL_INFO_DRAM_USAGE - Retrieve the dram usage inside the device and of the + * specific context. This is relevant only for GOYA device. + * HL_INFO_HW_IDLE - Retrieve information about the idle status of each + * internal engine. + * HL_INFO_DEVICE_STATUS - Retrieve the device's status. This opcode doesn't + * require an open context. + */ #define HL_INFO_HW_IP_INFO 0 #define HL_INFO_HW_EVENTS 1 #define HL_INFO_DRAM_USAGE 2 -- cgit v1.2.3-59-g8ed1b From 4fd2cb15cd0894520422b3e7d9cf68b7e6548b13 Mon Sep 17 00:00:00 2001 From: Dotan Barak Date: Mon, 12 Aug 2019 10:23:33 +0300 Subject: habanalabs: explicitly set the queue-id enumerated numbers When looking at kernel log messages and when debugging user applications, we only see the queue id. This patch explicitly set the queue id in the queue enumeration which will be helpful for finding the queue name when we have its id. Signed-off-by: Dotan Barak Reviewed-by: Oded Gabbay Reviewed-by: Greg Kroah-Hartman Signed-off-by: Oded Gabbay --- include/uapi/misc/habanalabs.h | 28 ++++++++++++++-------------- 1 file changed, 14 insertions(+), 14 deletions(-) (limited to 'include') diff --git a/include/uapi/misc/habanalabs.h b/include/uapi/misc/habanalabs.h index a5a1d0e7ec82..6cf50177cd21 100644 --- a/include/uapi/misc/habanalabs.h +++ b/include/uapi/misc/habanalabs.h @@ -28,20 +28,20 @@ enum goya_queue_id { GOYA_QUEUE_ID_DMA_0 = 0, - GOYA_QUEUE_ID_DMA_1, - GOYA_QUEUE_ID_DMA_2, - GOYA_QUEUE_ID_DMA_3, - GOYA_QUEUE_ID_DMA_4, - GOYA_QUEUE_ID_CPU_PQ, - GOYA_QUEUE_ID_MME, /* Internal queues start here */ - GOYA_QUEUE_ID_TPC0, - GOYA_QUEUE_ID_TPC1, - GOYA_QUEUE_ID_TPC2, - GOYA_QUEUE_ID_TPC3, - GOYA_QUEUE_ID_TPC4, - GOYA_QUEUE_ID_TPC5, - GOYA_QUEUE_ID_TPC6, - GOYA_QUEUE_ID_TPC7, + GOYA_QUEUE_ID_DMA_1 = 1, + GOYA_QUEUE_ID_DMA_2 = 2, + GOYA_QUEUE_ID_DMA_3 = 3, + GOYA_QUEUE_ID_DMA_4 = 4, + GOYA_QUEUE_ID_CPU_PQ = 5, + GOYA_QUEUE_ID_MME = 6, /* Internal queues start here */ + GOYA_QUEUE_ID_TPC0 = 7, + GOYA_QUEUE_ID_TPC1 = 8, + GOYA_QUEUE_ID_TPC2 = 9, + GOYA_QUEUE_ID_TPC3 = 10, + GOYA_QUEUE_ID_TPC4 = 11, + GOYA_QUEUE_ID_TPC5 = 12, + GOYA_QUEUE_ID_TPC6 = 13, + GOYA_QUEUE_ID_TPC7 = 14, GOYA_QUEUE_ID_SIZE }; -- cgit v1.2.3-59-g8ed1b From 413cf576fd50297429e967b050d63067c997645c Mon Sep 17 00:00:00 2001 From: Tomer Tayar Date: Tue, 27 Aug 2019 16:14:18 +0000 Subject: habanalabs: Make the Coresight timestamp perpetual The Coresight timestamp is enabled for a specific debug session using the HL_DEBUG_OP_TIMESTAMP opcode of the debug IOCTL. In order to have a perpetual timestamp that would be comparable between various debug sessions, this patch moves the timestamp enablement to be part of the HW initialization. The HL_DEBUG_OP_TIMESTAMP opcode turns to be deprecated and shouldn't be used. Old user-space that will call it won't see any change in the behavior of the debug session. Signed-off-by: Tomer Tayar Reviewed-by: Oded Gabbay Signed-off-by: Oded Gabbay --- drivers/misc/habanalabs/goya/goya.c | 23 +++++++++++++++++++++++ drivers/misc/habanalabs/goya/goya_coresight.c | 17 ++--------------- include/uapi/misc/habanalabs.h | 2 +- 3 files changed, 26 insertions(+), 16 deletions(-) (limited to 'include') diff --git a/drivers/misc/habanalabs/goya/goya.c b/drivers/misc/habanalabs/goya/goya.c index de275cb3bb98..0dd0b4429fee 100644 --- a/drivers/misc/habanalabs/goya/goya.c +++ b/drivers/misc/habanalabs/goya/goya.c @@ -2062,6 +2062,25 @@ static void goya_disable_msix(struct hl_device *hdev) goya->hw_cap_initialized &= ~HW_CAP_MSIX; } +static void goya_enable_timestamp(struct hl_device *hdev) +{ + /* Disable the timestamp counter */ + WREG32(mmPSOC_TIMESTAMP_BASE - CFG_BASE, 0); + + /* Zero the lower/upper parts of the 64-bit counter */ + WREG32(mmPSOC_TIMESTAMP_BASE - CFG_BASE + 0xC, 0); + WREG32(mmPSOC_TIMESTAMP_BASE - CFG_BASE + 0x8, 0); + + /* Enable the counter */ + WREG32(mmPSOC_TIMESTAMP_BASE - CFG_BASE, 1); +} + +static void goya_disable_timestamp(struct hl_device *hdev) +{ + /* Disable the timestamp counter */ + WREG32(mmPSOC_TIMESTAMP_BASE - CFG_BASE, 0); +} + static void goya_halt_engines(struct hl_device *hdev, bool hard_reset) { u32 wait_timeout_ms, cpu_timeout_ms; @@ -2102,6 +2121,8 @@ static void goya_halt_engines(struct hl_device *hdev, bool hard_reset) goya_disable_external_queues(hdev); goya_disable_internal_queues(hdev); + goya_disable_timestamp(hdev); + if (hard_reset) { goya_disable_msix(hdev); goya_mmu_remove_device_cpu_mappings(hdev); @@ -2504,6 +2525,8 @@ static int goya_hw_init(struct hl_device *hdev) goya_init_tpc_qmans(hdev); + goya_enable_timestamp(hdev); + /* MSI-X must be enabled before CPU queues are initialized */ rc = goya_enable_msix(hdev); if (rc) diff --git a/drivers/misc/habanalabs/goya/goya_coresight.c b/drivers/misc/habanalabs/goya/goya_coresight.c index 3d77b1c20336..b4d406af1bed 100644 --- a/drivers/misc/habanalabs/goya/goya_coresight.c +++ b/drivers/misc/habanalabs/goya/goya_coresight.c @@ -636,24 +636,11 @@ static int goya_config_spmu(struct hl_device *hdev, return 0; } -static int goya_config_timestamp(struct hl_device *hdev, - struct hl_debug_params *params) -{ - WREG32(mmPSOC_TIMESTAMP_BASE - CFG_BASE, 0); - if (params->enable) { - WREG32(mmPSOC_TIMESTAMP_BASE - CFG_BASE + 0xC, 0); - WREG32(mmPSOC_TIMESTAMP_BASE - CFG_BASE + 0x8, 0); - WREG32(mmPSOC_TIMESTAMP_BASE - CFG_BASE, 1); - } - - return 0; -} - int goya_debug_coresight(struct hl_device *hdev, void *data) { struct hl_debug_params *params = data; u32 val; - int rc; + int rc = 0; switch (params->op) { case HL_DEBUG_OP_STM: @@ -675,7 +662,7 @@ int goya_debug_coresight(struct hl_device *hdev, void *data) rc = goya_config_spmu(hdev, params); break; case HL_DEBUG_OP_TIMESTAMP: - rc = goya_config_timestamp(hdev, params); + /* Do nothing as this opcode is deprecated */ break; default: diff --git a/include/uapi/misc/habanalabs.h b/include/uapi/misc/habanalabs.h index 6cf50177cd21..266bf85056d4 100644 --- a/include/uapi/misc/habanalabs.h +++ b/include/uapi/misc/habanalabs.h @@ -451,7 +451,7 @@ struct hl_debug_params_spmu { #define HL_DEBUG_OP_BMON 4 /* Opcode for SPMU component */ #define HL_DEBUG_OP_SPMU 5 -/* Opcode for timestamp */ +/* Opcode for timestamp (deprecated) */ #define HL_DEBUG_OP_TIMESTAMP 6 /* Opcode for setting the device into or out of debug mode. The enable * variable should be 1 for enabling debug mode and 0 for disabling it -- cgit v1.2.3-59-g8ed1b From 75b3cb2bb080372d043e8f0c0aeae8f52461136b Mon Sep 17 00:00:00 2001 From: Oded Gabbay Date: Wed, 28 Aug 2019 17:32:04 +0300 Subject: habanalabs: add uapi to retrieve device utilization Users and sysadmins usually want to know what is the device utilization as a level 0 indication if they are efficiently using the device. Add a new opcode to the INFO IOCTL that will return the device utilization over the last period of 100-1000ms. The return value is 0-100, representing as percentage the total utilization rate. Signed-off-by: Oded Gabbay Reviewed-by: Omer Shpigelman --- drivers/misc/habanalabs/command_submission.c | 20 ++++- drivers/misc/habanalabs/device.c | 116 ++++++++++++++++++++++++++- drivers/misc/habanalabs/habanalabs.h | 23 +++++- drivers/misc/habanalabs/habanalabs_ioctl.c | 27 +++++++ drivers/misc/habanalabs/hw_queue.c | 8 +- drivers/misc/habanalabs/include/goya/goya.h | 2 + include/uapi/misc/habanalabs.h | 49 +++++++---- 7 files changed, 222 insertions(+), 23 deletions(-) (limited to 'include') diff --git a/drivers/misc/habanalabs/command_submission.c b/drivers/misc/habanalabs/command_submission.c index e4dd3e83df8b..4777ec4c2b55 100644 --- a/drivers/misc/habanalabs/command_submission.c +++ b/drivers/misc/habanalabs/command_submission.c @@ -178,11 +178,23 @@ static void cs_do_release(struct kref *ref) /* We also need to update CI for internal queues */ if (cs->submitted) { - int cs_cnt = atomic_dec_return(&hdev->cs_active_cnt); + hdev->asic_funcs->hw_queues_lock(hdev); - WARN_ONCE((cs_cnt < 0), - "hl%d: error in CS active cnt %d\n", - hdev->id, cs_cnt); + hdev->cs_active_cnt--; + if (!hdev->cs_active_cnt) { + struct hl_device_idle_busy_ts *ts; + + ts = &hdev->idle_busy_ts_arr[hdev->idle_busy_ts_idx++]; + ts->busy_to_idle_ts = ktime_get(); + + if (hdev->idle_busy_ts_idx == HL_IDLE_BUSY_TS_ARR_SIZE) + hdev->idle_busy_ts_idx = 0; + } else if (hdev->cs_active_cnt < 0) { + dev_crit(hdev->dev, "CS active cnt %d is negative\n", + hdev->cs_active_cnt); + } + + hdev->asic_funcs->hw_queues_unlock(hdev); hl_int_hw_queue_update_ci(cs); diff --git a/drivers/misc/habanalabs/device.c b/drivers/misc/habanalabs/device.c index 41c3ddbca351..cebdceb72298 100644 --- a/drivers/misc/habanalabs/device.c +++ b/drivers/misc/habanalabs/device.c @@ -293,6 +293,14 @@ static int device_early_init(struct hl_device *hdev) goto free_eq_wq; } + hdev->idle_busy_ts_arr = kmalloc_array(HL_IDLE_BUSY_TS_ARR_SIZE, + sizeof(struct hl_device_idle_busy_ts), + (GFP_KERNEL | __GFP_ZERO)); + if (!hdev->idle_busy_ts_arr) { + rc = -ENOMEM; + goto free_chip_info; + } + hl_cb_mgr_init(&hdev->kernel_cb_mgr); mutex_init(&hdev->send_cpu_message_lock); @@ -303,10 +311,11 @@ static int device_early_init(struct hl_device *hdev) INIT_LIST_HEAD(&hdev->fpriv_list); mutex_init(&hdev->fpriv_list_lock); atomic_set(&hdev->in_reset, 0); - atomic_set(&hdev->cs_active_cnt, 0); return 0; +free_chip_info: + kfree(hdev->hl_chip_info); free_eq_wq: destroy_workqueue(hdev->eq_wq); free_cq_wq: @@ -336,6 +345,7 @@ static void device_early_fini(struct hl_device *hdev) hl_cb_mgr_fini(hdev, &hdev->kernel_cb_mgr); + kfree(hdev->idle_busy_ts_arr); kfree(hdev->hl_chip_info); destroy_workqueue(hdev->eq_wq); @@ -451,6 +461,102 @@ static void device_late_fini(struct hl_device *hdev) hdev->late_init_done = false; } +uint32_t hl_device_utilization(struct hl_device *hdev, uint32_t period_ms) +{ + struct hl_device_idle_busy_ts *ts; + ktime_t zero_ktime, curr = ktime_get(); + u32 overlap_cnt = 0, last_index = hdev->idle_busy_ts_idx; + s64 period_us, last_start_us, last_end_us, last_busy_time_us, + total_busy_time_us = 0, total_busy_time_ms; + + zero_ktime = ktime_set(0, 0); + period_us = period_ms * USEC_PER_MSEC; + ts = &hdev->idle_busy_ts_arr[last_index]; + + /* check case that device is currently in idle */ + if (!ktime_compare(ts->busy_to_idle_ts, zero_ktime) && + !ktime_compare(ts->idle_to_busy_ts, zero_ktime)) { + + last_index--; + /* Handle case idle_busy_ts_idx was 0 */ + if (last_index > HL_IDLE_BUSY_TS_ARR_SIZE) + last_index = HL_IDLE_BUSY_TS_ARR_SIZE - 1; + + ts = &hdev->idle_busy_ts_arr[last_index]; + } + + while (overlap_cnt < HL_IDLE_BUSY_TS_ARR_SIZE) { + /* Check if we are in last sample case. i.e. if the sample + * begun before the sampling period. This could be a real + * sample or 0 so need to handle both cases + */ + last_start_us = ktime_to_us( + ktime_sub(curr, ts->idle_to_busy_ts)); + + if (last_start_us > period_us) { + + /* First check two cases: + * 1. If the device is currently busy + * 2. If the device was idle during the whole sampling + * period + */ + + if (!ktime_compare(ts->busy_to_idle_ts, zero_ktime)) { + /* Check if the device is currently busy */ + if (ktime_compare(ts->idle_to_busy_ts, + zero_ktime)) + return 100; + + /* We either didn't have any activity or we + * reached an entry which is 0. Either way, + * exit and return what was accumulated so far + */ + break; + } + + /* If sample has finished, check it is relevant */ + last_end_us = ktime_to_us( + ktime_sub(curr, ts->busy_to_idle_ts)); + + if (last_end_us > period_us) + break; + + /* It is relevant so add it but with adjustment */ + last_busy_time_us = ktime_to_us( + ktime_sub(ts->busy_to_idle_ts, + ts->idle_to_busy_ts)); + total_busy_time_us += last_busy_time_us - + (last_start_us - period_us); + break; + } + + /* Check if the sample is finished or still open */ + if (ktime_compare(ts->busy_to_idle_ts, zero_ktime)) + last_busy_time_us = ktime_to_us( + ktime_sub(ts->busy_to_idle_ts, + ts->idle_to_busy_ts)); + else + last_busy_time_us = ktime_to_us( + ktime_sub(curr, ts->idle_to_busy_ts)); + + total_busy_time_us += last_busy_time_us; + + last_index--; + /* Handle case idle_busy_ts_idx was 0 */ + if (last_index > HL_IDLE_BUSY_TS_ARR_SIZE) + last_index = HL_IDLE_BUSY_TS_ARR_SIZE - 1; + + ts = &hdev->idle_busy_ts_arr[last_index]; + + overlap_cnt++; + } + + total_busy_time_ms = DIV_ROUND_UP_ULL(total_busy_time_us, + USEC_PER_MSEC); + + return DIV_ROUND_UP_ULL(total_busy_time_ms * 100, period_ms); +} + /* * hl_device_set_frequency - set the frequency of the device * @@ -808,6 +914,14 @@ again: for (i = 0 ; i < hdev->asic_prop.completion_queues_count ; i++) hl_cq_reset(hdev, &hdev->completion_queue[i]); + hdev->idle_busy_ts_idx = 0; + hdev->idle_busy_ts_arr[0].busy_to_idle_ts = ktime_set(0, 0); + hdev->idle_busy_ts_arr[0].idle_to_busy_ts = ktime_set(0, 0); + + if (hdev->cs_active_cnt) + dev_crit(hdev->dev, "CS active cnt %d is not 0 during reset\n", + hdev->cs_active_cnt); + mutex_lock(&hdev->fpriv_list_lock); /* Make sure the context switch phase will run again */ diff --git a/drivers/misc/habanalabs/habanalabs.h b/drivers/misc/habanalabs/habanalabs.h index a4d929f5bad8..23b86b7f9732 100644 --- a/drivers/misc/habanalabs/habanalabs.h +++ b/drivers/misc/habanalabs/habanalabs.h @@ -45,6 +45,8 @@ /* MUST BE POWER OF 2 and larger than 1 */ #define HL_MAX_PENDING_CS 64 +#define HL_IDLE_BUSY_TS_ARR_SIZE 4096 + /* Memory */ #define MEM_HASH_TABLE_BITS 7 /* 1 << 7 buckets */ @@ -1156,6 +1158,16 @@ struct hl_device_reset_work { struct hl_device *hdev; }; +/** + * struct hl_device_idle_busy_ts - used for calculating device utilization rate. + * @idle_to_busy_ts: timestamp where device changed from idle to busy. + * @busy_to_idle_ts: timestamp where device changed from busy to idle. + */ +struct hl_device_idle_busy_ts { + ktime_t idle_to_busy_ts; + ktime_t busy_to_idle_ts; +}; + /** * struct hl_device - habanalabs device structure. * @pdev: pointer to PCI device, can be NULL in case of simulator device. @@ -1203,19 +1215,22 @@ struct hl_device_reset_work { * when a user opens the device * @fpriv_list_lock: protects the fpriv_list * @compute_ctx: current compute context executing. + * @idle_busy_ts_arr: array to hold time stamps of transitions from idle to busy + * and vice-versa * @dram_used_mem: current DRAM memory consumption. * @timeout_jiffies: device CS timeout value. * @max_power: the max power of the device, as configured by the sysadmin. This * value is saved so in case of hard-reset, KMD will restore this * value and update the F/W after the re-initialization * @in_reset: is device in reset flow. + * @curr_pll_profile: current PLL profile. * @cs_active_cnt: number of active command submissions on this device (active * means already in H/W queues) - * @curr_pll_profile: current PLL profile. * @major: habanalabs KMD major. * @high_pll: high PLL profile frequency. * @soft_reset_cnt: number of soft reset since KMD loading. * @hard_reset_cnt: number of hard reset since KMD loading. + * @idle_busy_ts_idx: index of current entry in idle_busy_ts_arr * @id: device minor. * @id_control: minor of the control device * @disabled: is device disabled. @@ -1285,16 +1300,19 @@ struct hl_device { struct hl_ctx *compute_ctx; + struct hl_device_idle_busy_ts *idle_busy_ts_arr; + atomic64_t dram_used_mem; u64 timeout_jiffies; u64 max_power; atomic_t in_reset; - atomic_t cs_active_cnt; enum hl_pll_frequency curr_pll_profile; + int cs_active_cnt; u32 major; u32 high_pll; u32 soft_reset_cnt; u32 hard_reset_cnt; + u32 idle_busy_ts_idx; u16 id; u16 id_control; u8 disabled; @@ -1457,6 +1475,7 @@ int hl_device_reset(struct hl_device *hdev, bool hard_reset, void hl_hpriv_get(struct hl_fpriv *hpriv); void hl_hpriv_put(struct hl_fpriv *hpriv); int hl_device_set_frequency(struct hl_device *hdev, enum hl_pll_frequency freq); +uint32_t hl_device_utilization(struct hl_device *hdev, uint32_t period_ms); int hl_build_hwmon_channel_info(struct hl_device *hdev, struct armcp_sensor *sensors_arr); diff --git a/drivers/misc/habanalabs/habanalabs_ioctl.c b/drivers/misc/habanalabs/habanalabs_ioctl.c index 8fbca2c0f44b..f958568f7996 100644 --- a/drivers/misc/habanalabs/habanalabs_ioctl.c +++ b/drivers/misc/habanalabs/habanalabs_ioctl.c @@ -197,6 +197,29 @@ out: return rc; } +static int device_utilization(struct hl_device *hdev, struct hl_info_args *args) +{ + struct hl_info_device_utilization device_util = {0}; + u32 max_size = args->return_size; + void __user *out = (void __user *) (uintptr_t) args->return_pointer; + + if ((!max_size) || (!out)) + return -EINVAL; + + if ((args->period_ms < 100) || (args->period_ms > 1000) || + (args->period_ms % 100)) { + dev_err(hdev->dev, + "period %u must be between 100 - 1000 and must be divisible by 100\n", + args->period_ms); + return -EINVAL; + } + + device_util.utilization = hl_device_utilization(hdev, args->period_ms); + + return copy_to_user(out, &device_util, + min((size_t) max_size, sizeof(device_util))) ? -EFAULT : 0; +} + static int _hl_info_ioctl(struct hl_fpriv *hpriv, void *data, struct device *dev) { @@ -239,6 +262,10 @@ static int _hl_info_ioctl(struct hl_fpriv *hpriv, void *data, rc = hw_idle(hdev, args); break; + case HL_INFO_DEVICE_UTILIZATION: + rc = device_utilization(hdev, args); + break; + default: dev_err(dev, "Invalid request %d\n", args->op); rc = -ENOTTY; diff --git a/drivers/misc/habanalabs/hw_queue.c b/drivers/misc/habanalabs/hw_queue.c index 696cf7d206c6..55b383b2a116 100644 --- a/drivers/misc/habanalabs/hw_queue.c +++ b/drivers/misc/habanalabs/hw_queue.c @@ -364,7 +364,13 @@ int hl_hw_queue_schedule_cs(struct hl_cs *cs) spin_unlock(&hdev->hw_queues_mirror_lock); } - atomic_inc(&hdev->cs_active_cnt); + if (!hdev->cs_active_cnt++) { + struct hl_device_idle_busy_ts *ts; + + ts = &hdev->idle_busy_ts_arr[hdev->idle_busy_ts_idx]; + ts->busy_to_idle_ts = ktime_set(0, 0); + ts->idle_to_busy_ts = ktime_get(); + } list_for_each_entry_safe(job, tmp, &cs->job_list, cs_node) if (job->ext_queue) diff --git a/drivers/misc/habanalabs/include/goya/goya.h b/drivers/misc/habanalabs/include/goya/goya.h index 3f02a52ba4ce..43d241891e45 100644 --- a/drivers/misc/habanalabs/include/goya/goya.h +++ b/drivers/misc/habanalabs/include/goya/goya.h @@ -38,4 +38,6 @@ #define TPC_MAX_NUM 8 +#define MME_MAX_NUM 1 + #endif /* GOYA_H */ diff --git a/include/uapi/misc/habanalabs.h b/include/uapi/misc/habanalabs.h index 266bf85056d4..73ee212d7fa6 100644 --- a/include/uapi/misc/habanalabs.h +++ b/include/uapi/misc/habanalabs.h @@ -77,22 +77,29 @@ enum hl_device_status { /* Opcode for management ioctl * - * HW_IP_INFO - Receive information about different IP blocks in the - * device. - * HL_INFO_HW_EVENTS - Receive an array describing how many times each event - * occurred since the last hard reset. - * HL_INFO_DRAM_USAGE - Retrieve the dram usage inside the device and of the - * specific context. This is relevant only for GOYA device. - * HL_INFO_HW_IDLE - Retrieve information about the idle status of each - * internal engine. + * HW_IP_INFO - Receive information about different IP blocks in the + * device. + * HL_INFO_HW_EVENTS - Receive an array describing how many times each event + * occurred since the last hard reset. + * HL_INFO_DRAM_USAGE - Retrieve the dram usage inside the device and of the + * specific context. This is relevant only for devices + * where the dram is managed by the kernel driver + * HL_INFO_HW_IDLE - Retrieve information about the idle status of each + * internal engine. * HL_INFO_DEVICE_STATUS - Retrieve the device's status. This opcode doesn't * require an open context. + * HL_INFO_DEVICE_UTILIZATION - Retrieve the total utilization of the device + * over the last period specified by the user. + * The period can be between 100ms to 1s, in + * resolution of 100ms. The return value is a + * percentage of the utilization rate. */ -#define HL_INFO_HW_IP_INFO 0 -#define HL_INFO_HW_EVENTS 1 -#define HL_INFO_DRAM_USAGE 2 -#define HL_INFO_HW_IDLE 3 -#define HL_INFO_DEVICE_STATUS 4 +#define HL_INFO_HW_IP_INFO 0 +#define HL_INFO_HW_EVENTS 1 +#define HL_INFO_DRAM_USAGE 2 +#define HL_INFO_HW_IDLE 3 +#define HL_INFO_DEVICE_STATUS 4 +#define HL_INFO_DEVICE_UTILIZATION 6 #define HL_INFO_VERSION_MAX_LEN 128 @@ -134,6 +141,11 @@ struct hl_info_device_status { __u32 pad; }; +struct hl_info_device_utilization { + __u32 utilization; + __u32 pad; +}; + struct hl_info_args { /* Location of relevant struct in userspace */ __u64 return_pointer; @@ -149,8 +161,15 @@ struct hl_info_args { /* HL_INFO_* */ __u32 op; - /* Context ID - Currently not in use */ - __u32 ctx_id; + union { + /* Context ID - Currently not in use */ + __u32 ctx_id; + /* Period value for utilization rate (100ms - 1000ms, in 100ms + * resolution. + */ + __u32 period_ms; + }; + __u32 pad; }; -- cgit v1.2.3-59-g8ed1b From e9730763a21a5441d46511f124d703d76a5ef6e6 Mon Sep 17 00:00:00 2001 From: Oded Gabbay Date: Wed, 28 Aug 2019 21:51:52 +0300 Subject: habanalabs: add uapi to retrieve aggregate H/W events Add a new opcode to INFO IOCTL to retrieve aggregate H/W events. i.e. the events counters are NOT cleared upon device reset, but count from the loading of the driver. Add the code to support it in the device event handling function. Signed-off-by: Oded Gabbay Reviewed-by: Omer Shpigelman --- drivers/misc/habanalabs/goya/goya.c | 9 +++++++-- drivers/misc/habanalabs/goya/goyaP.h | 3 ++- drivers/misc/habanalabs/habanalabs.h | 3 ++- drivers/misc/habanalabs/habanalabs_ioctl.c | 11 ++++++++--- include/uapi/misc/habanalabs.h | 3 +++ 5 files changed, 22 insertions(+), 7 deletions(-) (limited to 'include') diff --git a/drivers/misc/habanalabs/goya/goya.c b/drivers/misc/habanalabs/goya/goya.c index 0dd0b4429fee..1267ec75b19f 100644 --- a/drivers/misc/habanalabs/goya/goya.c +++ b/drivers/misc/habanalabs/goya/goya.c @@ -4469,6 +4469,7 @@ void goya_handle_eqe(struct hl_device *hdev, struct hl_eq_entry *eq_entry) struct goya_device *goya = hdev->asic_specific; goya->events_stat[event_type]++; + goya->events_stat_aggregate[event_type]++; switch (event_type) { case GOYA_ASYNC_EVENT_ID_PCIE_IF: @@ -4550,12 +4551,16 @@ void goya_handle_eqe(struct hl_device *hdev, struct hl_eq_entry *eq_entry) } } -void *goya_get_events_stat(struct hl_device *hdev, u32 *size) +void *goya_get_events_stat(struct hl_device *hdev, bool aggregate, u32 *size) { struct goya_device *goya = hdev->asic_specific; - *size = (u32) sizeof(goya->events_stat); + if (aggregate) { + *size = (u32) sizeof(goya->events_stat_aggregate); + return goya->events_stat_aggregate; + } + *size = (u32) sizeof(goya->events_stat); return goya->events_stat; } diff --git a/drivers/misc/habanalabs/goya/goyaP.h b/drivers/misc/habanalabs/goya/goyaP.h index d7f48c9c41cd..f830cfd5c04d 100644 --- a/drivers/misc/habanalabs/goya/goyaP.h +++ b/drivers/misc/habanalabs/goya/goyaP.h @@ -162,6 +162,7 @@ struct goya_device { u64 ddr_bar_cur_addr; u32 events_stat[GOYA_ASYNC_EVENT_ID_SIZE]; + u32 events_stat_aggregate[GOYA_ASYNC_EVENT_ID_SIZE]; u32 hw_cap_initialized; u8 device_cpu_mmu_mappings_done; }; @@ -215,7 +216,7 @@ int goya_suspend(struct hl_device *hdev); int goya_resume(struct hl_device *hdev); void goya_handle_eqe(struct hl_device *hdev, struct hl_eq_entry *eq_entry); -void *goya_get_events_stat(struct hl_device *hdev, u32 *size); +void *goya_get_events_stat(struct hl_device *hdev, bool aggregate, u32 *size); void goya_add_end_of_cb_packets(struct hl_device *hdev, u64 kernel_address, u32 len, u64 cq_addr, u32 cq_val, u32 msix_vec); diff --git a/drivers/misc/habanalabs/habanalabs.h b/drivers/misc/habanalabs/habanalabs.h index 23b86b7f9732..aa7aaa710f12 100644 --- a/drivers/misc/habanalabs/habanalabs.h +++ b/drivers/misc/habanalabs/habanalabs.h @@ -558,7 +558,8 @@ struct hl_asic_funcs { struct hl_eq_entry *eq_entry); void (*set_pll_profile)(struct hl_device *hdev, enum hl_pll_frequency freq); - void* (*get_events_stat)(struct hl_device *hdev, u32 *size); + void* (*get_events_stat)(struct hl_device *hdev, bool aggregate, + u32 *size); u64 (*read_pte)(struct hl_device *hdev, u64 addr); void (*write_pte)(struct hl_device *hdev, u64 addr, u64 val); void (*mmu_invalidate_cache)(struct hl_device *hdev, bool is_hard); diff --git a/drivers/misc/habanalabs/habanalabs_ioctl.c b/drivers/misc/habanalabs/habanalabs_ioctl.c index f958568f7996..66d9c710073c 100644 --- a/drivers/misc/habanalabs/habanalabs_ioctl.c +++ b/drivers/misc/habanalabs/habanalabs_ioctl.c @@ -75,7 +75,8 @@ static int hw_ip_info(struct hl_device *hdev, struct hl_info_args *args) min((size_t)size, sizeof(hw_ip))) ? -EFAULT : 0; } -static int hw_events_info(struct hl_device *hdev, struct hl_info_args *args) +static int hw_events_info(struct hl_device *hdev, bool aggregate, + struct hl_info_args *args) { u32 size, max_size = args->return_size; void __user *out = (void __user *) (uintptr_t) args->return_pointer; @@ -84,7 +85,7 @@ static int hw_events_info(struct hl_device *hdev, struct hl_info_args *args) if ((!max_size) || (!out)) return -EINVAL; - arr = hdev->asic_funcs->get_events_stat(hdev, &size); + arr = hdev->asic_funcs->get_events_stat(hdev, aggregate, &size); return copy_to_user(out, arr, min(max_size, size)) ? -EFAULT : 0; } @@ -251,7 +252,7 @@ static int _hl_info_ioctl(struct hl_fpriv *hpriv, void *data, switch (args->op) { case HL_INFO_HW_EVENTS: - rc = hw_events_info(hdev, args); + rc = hw_events_info(hdev, false, args); break; case HL_INFO_DRAM_USAGE: @@ -266,6 +267,10 @@ static int _hl_info_ioctl(struct hl_fpriv *hpriv, void *data, rc = device_utilization(hdev, args); break; + case HL_INFO_HW_EVENTS_AGGREGATE: + rc = hw_events_info(hdev, true, args); + break; + default: dev_err(dev, "Invalid request %d\n", args->op); rc = -ENOTTY; diff --git a/include/uapi/misc/habanalabs.h b/include/uapi/misc/habanalabs.h index 73ee212d7fa6..19f8039db2ea 100644 --- a/include/uapi/misc/habanalabs.h +++ b/include/uapi/misc/habanalabs.h @@ -93,6 +93,8 @@ enum hl_device_status { * The period can be between 100ms to 1s, in * resolution of 100ms. The return value is a * percentage of the utilization rate. + * HL_INFO_HW_EVENTS_AGGREGATE - Receive an array describing how many times each + * event occurred since the driver was loaded. */ #define HL_INFO_HW_IP_INFO 0 #define HL_INFO_HW_EVENTS 1 @@ -100,6 +102,7 @@ enum hl_device_status { #define HL_INFO_HW_IDLE 3 #define HL_INFO_DEVICE_STATUS 4 #define HL_INFO_DEVICE_UTILIZATION 6 +#define HL_INFO_HW_EVENTS_AGGREGATE 7 #define HL_INFO_VERSION_MAX_LEN 128 -- cgit v1.2.3-59-g8ed1b From 4c172bbfaa4e1aa26dab58781301902c7b3e4ebc Mon Sep 17 00:00:00 2001 From: Oded Gabbay Date: Fri, 30 Aug 2019 16:59:33 +0300 Subject: habanalabs: stop using the acronym KMD We want to stop using the acronym KMD. Therefore, replace all locations (except for register names we can't modify) where KMD is written to other terms such as "Linux kernel driver" or "Host kernel driver", etc. Signed-off-by: Oded Gabbay Reviewed-by: Omer Shpigelman --- drivers/misc/habanalabs/asid.c | 2 +- drivers/misc/habanalabs/command_buffer.c | 3 +- drivers/misc/habanalabs/command_submission.c | 5 +- drivers/misc/habanalabs/context.c | 2 +- drivers/misc/habanalabs/goya/goya.c | 22 ++++----- drivers/misc/habanalabs/goya/goyaP.h | 14 +++--- drivers/misc/habanalabs/habanalabs.h | 32 ++++++------- drivers/misc/habanalabs/include/armcp_if.h | 70 ++++++++++++++-------------- include/uapi/misc/habanalabs.h | 22 ++++----- 9 files changed, 88 insertions(+), 84 deletions(-) (limited to 'include') diff --git a/drivers/misc/habanalabs/asid.c b/drivers/misc/habanalabs/asid.c index 2c01461701a3..a2fdf31cf27c 100644 --- a/drivers/misc/habanalabs/asid.c +++ b/drivers/misc/habanalabs/asid.c @@ -18,7 +18,7 @@ int hl_asid_init(struct hl_device *hdev) mutex_init(&hdev->asid_mutex); - /* ASID 0 is reserved for KMD and device CPU */ + /* ASID 0 is reserved for the kernel driver and device CPU */ set_bit(0, hdev->asid_bitmap); return 0; diff --git a/drivers/misc/habanalabs/command_buffer.c b/drivers/misc/habanalabs/command_buffer.c index e495f44064fa..53fddbd8e693 100644 --- a/drivers/misc/habanalabs/command_buffer.c +++ b/drivers/misc/habanalabs/command_buffer.c @@ -397,7 +397,8 @@ struct hl_cb *hl_cb_kernel_create(struct hl_device *hdev, u32 cb_size) rc = hl_cb_create(hdev, &hdev->kernel_cb_mgr, cb_size, &cb_handle, HL_KERNEL_ASID_ID); if (rc) { - dev_err(hdev->dev, "Failed to allocate CB for KMD %d\n", rc); + dev_err(hdev->dev, + "Failed to allocate CB for the kernel driver %d\n", rc); return NULL; } diff --git a/drivers/misc/habanalabs/command_submission.c b/drivers/misc/habanalabs/command_submission.c index 4777ec4c2b55..a9ac045dcfde 100644 --- a/drivers/misc/habanalabs/command_submission.c +++ b/drivers/misc/habanalabs/command_submission.c @@ -409,8 +409,9 @@ static struct hl_cb *validate_queue_index(struct hl_device *hdev, return NULL; } - if (hw_queue_prop->kmd_only) { - dev_err(hdev->dev, "Queue index %d is restricted for KMD\n", + if (hw_queue_prop->driver_only) { + dev_err(hdev->dev, + "Queue index %d is restricted for the kernel driver\n", chunk->queue_index); return NULL; } else if (hw_queue_prop->type == QUEUE_TYPE_INT) { diff --git a/drivers/misc/habanalabs/context.c b/drivers/misc/habanalabs/context.c index bc0dec57a983..17db7b3dfb4c 100644 --- a/drivers/misc/habanalabs/context.c +++ b/drivers/misc/habanalabs/context.c @@ -128,7 +128,7 @@ int hl_ctx_init(struct hl_device *hdev, struct hl_ctx *ctx, bool is_kernel_ctx) ctx->thread_ctx_switch_wait_token = 0; if (is_kernel_ctx) { - ctx->asid = HL_KERNEL_ASID_ID; /* KMD gets ASID 0 */ + ctx->asid = HL_KERNEL_ASID_ID; /* Kernel driver gets ASID 0 */ rc = hl_mmu_ctx_init(ctx); if (rc) { dev_err(hdev->dev, "Failed to init mmu ctx module\n"); diff --git a/drivers/misc/habanalabs/goya/goya.c b/drivers/misc/habanalabs/goya/goya.c index c88c2fea97b9..6fba14b81f90 100644 --- a/drivers/misc/habanalabs/goya/goya.c +++ b/drivers/misc/habanalabs/goya/goya.c @@ -42,8 +42,8 @@ * PQ, CQ and CP are not secured. * PQ, CB and the data are on the SRAM/DRAM. * - * Since QMAN DMA is secured, KMD is parsing the DMA CB: - * - KMD checks DMA pointer + * Since QMAN DMA is secured, the driver is parsing the DMA CB: + * - checks DMA pointer * - WREG, MSG_PROT are not allowed. * - MSG_LONG/SHORT are allowed. * @@ -56,15 +56,15 @@ * QMAN DMA: PQ, CQ and CP are secured. * MMU is set to bypass on the Secure props register of the QMAN. * The reasons we don't enable MMU for PQ, CQ and CP are: - * - PQ entry is in kernel address space and KMD doesn't map it. + * - PQ entry is in kernel address space and the driver doesn't map it. * - CP writes to MSIX register and to kernel address space (completion * queue). * - * DMA is not secured but because CP is secured, KMD still needs to parse the - * CB, but doesn't need to check the DMA addresses. + * DMA is not secured but because CP is secured, the driver still needs to parse + * the CB, but doesn't need to check the DMA addresses. * - * For QMAN DMA 0, DMA is also secured because only KMD uses this DMA and KMD - * doesn't map memory in MMU. + * For QMAN DMA 0, DMA is also secured because only the driver uses this DMA and + * the driver doesn't map memory in MMU. * * QMAN TPC/MME: PQ, CQ and CP aren't secured (no change from MMU disabled mode) * @@ -336,18 +336,18 @@ void goya_get_fixed_properties(struct hl_device *hdev) for (i = 0 ; i < NUMBER_OF_EXT_HW_QUEUES ; i++) { prop->hw_queues_props[i].type = QUEUE_TYPE_EXT; - prop->hw_queues_props[i].kmd_only = 0; + prop->hw_queues_props[i].driver_only = 0; } for (; i < NUMBER_OF_EXT_HW_QUEUES + NUMBER_OF_CPU_HW_QUEUES ; i++) { prop->hw_queues_props[i].type = QUEUE_TYPE_CPU; - prop->hw_queues_props[i].kmd_only = 1; + prop->hw_queues_props[i].driver_only = 1; } for (; i < NUMBER_OF_EXT_HW_QUEUES + NUMBER_OF_CPU_HW_QUEUES + NUMBER_OF_INT_HW_QUEUES; i++) { prop->hw_queues_props[i].type = QUEUE_TYPE_INT; - prop->hw_queues_props[i].kmd_only = 0; + prop->hw_queues_props[i].driver_only = 0; } for (; i < HL_MAX_QUEUES; i++) @@ -2853,7 +2853,7 @@ static int goya_send_job_on_qman0(struct hl_device *hdev, struct hl_cs_job *job) if (!hdev->asic_funcs->is_device_idle(hdev, NULL, NULL)) { dev_err_ratelimited(hdev->dev, - "Can't send KMD job on QMAN0 because the device is not idle\n"); + "Can't send driver job on QMAN0 because the device is not idle\n"); return -EBUSY; } diff --git a/drivers/misc/habanalabs/goya/goyaP.h b/drivers/misc/habanalabs/goya/goyaP.h index 06da71e8d7ea..89b6574f8e4f 100644 --- a/drivers/misc/habanalabs/goya/goyaP.h +++ b/drivers/misc/habanalabs/goya/goyaP.h @@ -70,19 +70,19 @@ MMU_PAGE_TABLES_SIZE) #define MMU_CACHE_MNG_ADDR (MMU_DRAM_DEFAULT_PAGE_ADDR + \ MMU_DRAM_DEFAULT_PAGE_SIZE) -#define DRAM_KMD_END_ADDR (MMU_CACHE_MNG_ADDR + \ +#define DRAM_DRIVER_END_ADDR (MMU_CACHE_MNG_ADDR + \ MMU_CACHE_MNG_SIZE) #define DRAM_BASE_ADDR_USER 0x20000000 -#if (DRAM_KMD_END_ADDR > DRAM_BASE_ADDR_USER) -#error "KMD must reserve no more than 512MB" +#if (DRAM_DRIVER_END_ADDR > DRAM_BASE_ADDR_USER) +#error "Driver must reserve no more than 512MB" #endif /* - * SRAM Memory Map for KMD + * SRAM Memory Map for Driver * - * KMD occupies KMD_SRAM_SIZE bytes from the start of SRAM. It is used for + * Driver occupies DRIVER_SRAM_SIZE bytes from the start of SRAM. It is used for * MME/TPC QMANs * */ @@ -108,10 +108,10 @@ #define TPC7_QMAN_BASE_OFFSET (TPC6_QMAN_BASE_OFFSET + \ (TPC_QMAN_LENGTH * QMAN_PQ_ENTRY_SIZE)) -#define SRAM_KMD_RES_OFFSET (TPC7_QMAN_BASE_OFFSET + \ +#define SRAM_DRIVER_RES_OFFSET (TPC7_QMAN_BASE_OFFSET + \ (TPC_QMAN_LENGTH * QMAN_PQ_ENTRY_SIZE)) -#if (SRAM_KMD_RES_OFFSET >= GOYA_KMD_SRAM_RESERVED_SIZE_FROM_START) +#if (SRAM_DRIVER_RES_OFFSET >= GOYA_KMD_SRAM_RESERVED_SIZE_FROM_START) #error "MME/TPC QMANs SRAM space exceeds limit" #endif diff --git a/drivers/misc/habanalabs/habanalabs.h b/drivers/misc/habanalabs/habanalabs.h index aa7aaa710f12..c39e07d665c4 100644 --- a/drivers/misc/habanalabs/habanalabs.h +++ b/drivers/misc/habanalabs/habanalabs.h @@ -96,12 +96,12 @@ enum hl_queue_type { /** * struct hw_queue_properties - queue information. * @type: queue type. - * @kmd_only: true if only KMD is allowed to send a job to this queue, false - * otherwise. + * @driver_only: true if only the driver is allowed to send a job to this queue, + * false otherwise. */ struct hw_queue_properties { enum hl_queue_type type; - u8 kmd_only; + u8 driver_only; }; /** @@ -324,7 +324,7 @@ struct hl_cs_job; #define HL_EQ_LENGTH 64 #define HL_EQ_SIZE_IN_BYTES (HL_EQ_LENGTH * HL_EQ_ENTRY_SIZE) -/* KMD <-> ArmCP shared memory size */ +/* Host <-> ArmCP shared memory size */ #define HL_CPU_ACCESSIBLE_MEM_SIZE SZ_2M /** @@ -405,7 +405,7 @@ struct hl_cs_parser; /** * enum hl_pm_mng_profile - power management profile. - * @PM_AUTO: internal clock is set by KMD. + * @PM_AUTO: internal clock is set by the Linux driver. * @PM_MANUAL: internal clock is set by the user. * @PM_LAST: last power management type. */ @@ -613,7 +613,7 @@ struct hl_va_range { * descriptor (hl_vm_phys_pg_list or hl_userptr). * @mmu_phys_hash: holds a mapping from physical address to pgt_info structure. * @mmu_shadow_hash: holds a mapping from shadow address to pgt_info structure. - * @hpriv: pointer to the private (KMD) data of the process (fd). + * @hpriv: pointer to the private (Kernel Driver) data of the process (fd). * @hdev: pointer to the device structure. * @refcount: reference counter for the context. Context is released only when * this hits 0l. It is incremented on CS and CS_WAIT. @@ -1185,19 +1185,19 @@ struct hl_device_idle_busy_ts { * @completion_queue: array of hl_cq. * @cq_wq: work queue of completion queues for executing work in process context * @eq_wq: work queue of event queue for executing work in process context. - * @kernel_ctx: KMD context structure. + * @kernel_ctx: Kernel driver context structure. * @kernel_queues: array of hl_hw_queue. * @hw_queues_mirror_list: CS mirror list for TDR. * @hw_queues_mirror_lock: protects hw_queues_mirror_list. * @kernel_cb_mgr: command buffer manager for creating/destroying/handling CGs. * @event_queue: event queue for IRQ from ArmCP. * @dma_pool: DMA pool for small allocations. - * @cpu_accessible_dma_mem: KMD <-> ArmCP shared memory CPU address. - * @cpu_accessible_dma_address: KMD <-> ArmCP shared memory DMA address. - * @cpu_accessible_dma_pool: KMD <-> ArmCP shared memory pool. + * @cpu_accessible_dma_mem: Host <-> ArmCP shared memory CPU address. + * @cpu_accessible_dma_address: Host <-> ArmCP shared memory DMA address. + * @cpu_accessible_dma_pool: Host <-> ArmCP shared memory pool. * @asid_bitmap: holds used/available ASIDs. * @asid_mutex: protects asid_bitmap. - * @send_cpu_message_lock: enforces only one message in KMD <-> ArmCP queue. + * @send_cpu_message_lock: enforces only one message in Host <-> ArmCP queue. * @debug_lock: protects critical section of setting debug mode for device * @asic_prop: ASIC specific immutable properties. * @asic_funcs: ASIC specific functions. @@ -1221,16 +1221,16 @@ struct hl_device_idle_busy_ts { * @dram_used_mem: current DRAM memory consumption. * @timeout_jiffies: device CS timeout value. * @max_power: the max power of the device, as configured by the sysadmin. This - * value is saved so in case of hard-reset, KMD will restore this - * value and update the F/W after the re-initialization + * value is saved so in case of hard-reset, the driver will restore + * this value and update the F/W after the re-initialization * @in_reset: is device in reset flow. * @curr_pll_profile: current PLL profile. * @cs_active_cnt: number of active command submissions on this device (active * means already in H/W queues) - * @major: habanalabs KMD major. + * @major: habanalabs kernel driver major. * @high_pll: high PLL profile frequency. - * @soft_reset_cnt: number of soft reset since KMD loading. - * @hard_reset_cnt: number of hard reset since KMD loading. + * @soft_reset_cnt: number of soft reset since the driver was loaded. + * @hard_reset_cnt: number of hard reset since the driver was loaded. * @idle_busy_ts_idx: index of current entry in idle_busy_ts_arr * @id: device minor. * @id_control: minor of the control device diff --git a/drivers/misc/habanalabs/include/armcp_if.h b/drivers/misc/habanalabs/include/armcp_if.h index 5565bce60bc9..e4c6699a1868 100644 --- a/drivers/misc/habanalabs/include/armcp_if.h +++ b/drivers/misc/habanalabs/include/armcp_if.h @@ -41,33 +41,34 @@ enum pq_init_status { /* * ArmCP Primary Queue Packets * - * During normal operation, KMD needs to send various messages to ArmCP, - * usually either to SET some value into a H/W periphery or to GET the current - * value of some H/W periphery. For example, SET the frequency of MME/TPC and - * GET the value of the thermal sensor. - * - * These messages can be initiated either by the User application or by KMD - * itself, e.g. power management code. In either case, the communication from - * KMD to ArmCP will *always* be in synchronous mode, meaning that KMD will - * send a single message and poll until the message was acknowledged and the - * results are ready (if results are needed). - * - * This means that only a single message can be sent at a time and KMD must - * wait for its result before sending the next message. Having said that, - * because these are control messages which are sent in a relatively low + * During normal operation, the host's kernel driver needs to send various + * messages to ArmCP, usually either to SET some value into a H/W periphery or + * to GET the current value of some H/W periphery. For example, SET the + * frequency of MME/TPC and GET the value of the thermal sensor. + * + * These messages can be initiated either by the User application or by the + * host's driver itself, e.g. power management code. In either case, the + * communication from the host's driver to ArmCP will *always* be in + * synchronous mode, meaning that the host will send a single message and poll + * until the message was acknowledged and the results are ready (if results are + * needed). + * + * This means that only a single message can be sent at a time and the host's + * driver must wait for its result before sending the next message. Having said + * that, because these are control messages which are sent in a relatively low * frequency, this limitation seems acceptable. It's important to note that * in case of multiple devices, messages to different devices *can* be sent * at the same time. * * The message, inputs/outputs (if relevant) and fence object will be located - * on the device DDR at an address that will be determined by KMD. During - * device initialization phase, KMD will pass to ArmCP that address. Most of - * the message types will contain inputs/outputs inside the message itself. - * The common part of each message will contain the opcode of the message (its - * type) and a field representing a fence object. - * - * When KMD wishes to send a message to ArmCP, it will write the message - * contents to the device DDR, clear the fence object and then write the + * on the device DDR at an address that will be determined by the host's driver. + * During device initialization phase, the host will pass to ArmCP that address. + * Most of the message types will contain inputs/outputs inside the message + * itself. The common part of each message will contain the opcode of the + * message (its type) and a field representing a fence object. + * + * When the host's driver wishes to send a message to ArmCP, it will write the + * message contents to the device DDR, clear the fence object and then write the * value 484 to the mmGIC_DISTRIBUTOR__5_GICD_SETSPI_NSR register to issue * the 484 interrupt-id to the ARM core. * @@ -78,12 +79,13 @@ enum pq_init_status { * device DDR and then write to the fence object. If an error occurred, ArmCP * will fill the rc field with the right error code. * - * In the meantime, KMD will poll on the fence object. Once KMD sees that the - * fence object is signaled, it will read the results from the device DDR - * (if relevant) and resume the code execution in KMD. + * In the meantime, the host's driver will poll on the fence object. Once the + * host sees that the fence object is signaled, it will read the results from + * the device DDR (if relevant) and resume the code execution in the host's + * driver. * * To use QMAN packets, the opcode must be the QMAN opcode, shifted by 8 - * so the value being put by the KMD matches the value read by ArmCP + * so the value being put by the host's driver matches the value read by ArmCP * * Non-QMAN packets should be limited to values 1 through (2^8 - 1) * @@ -148,9 +150,9 @@ enum pq_init_status { * * ARMCP_PACKET_INFO_GET - * Fetch information from the device as specified in the packet's - * structure. KMD passes the max size it allows the ArmCP to write to - * the structure, to prevent data corruption in case of mismatched - * KMD/FW versions. + * structure. The host's driver passes the max size it allows the ArmCP to + * write to the structure, to prevent data corruption in case of + * mismatched driver/FW versions. * * ARMCP_PACKET_FLASH_PROGRAM_REMOVED - this packet was removed * @@ -183,9 +185,9 @@ enum pq_init_status { * ARMCP_PACKET_EEPROM_DATA_GET - * Get EEPROM data from the ArmCP kernel. The buffer is specified in the * addr field. The CPU will put the returned data size in the result - * field. In addition, KMD passes the max size it allows the ArmCP to - * write to the structure, to prevent data corruption in case of - * mismatched KMD/FW versions. + * field. In addition, the host's driver passes the max size it allows the + * ArmCP to write to the structure, to prevent data corruption in case of + * mismatched driver/FW versions. * */ @@ -231,7 +233,7 @@ struct armcp_packet { __le32 ctl; - __le32 fence; /* Signal to KMD that message is completed */ + __le32 fence; /* Signal to host that message is completed */ union { struct {/* For temperature/current/voltage/fan/pwm get/set */ @@ -320,7 +322,7 @@ struct armcp_sensor { }; /** - * struct armcp_info - host driver's necessary info from ArmCP. + * struct armcp_info - Info from ArmCP that is necessary to the host's driver * @sensors: available sensors description. * @kernel_version: ArmCP linux kernel version. * @reserved: reserved field. diff --git a/include/uapi/misc/habanalabs.h b/include/uapi/misc/habanalabs.h index 19f8039db2ea..39c4ea51a719 100644 --- a/include/uapi/misc/habanalabs.h +++ b/include/uapi/misc/habanalabs.h @@ -1,6 +1,6 @@ /* SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note * - * Copyright 2016-2018 HabanaLabs, Ltd. + * Copyright 2016-2019 HabanaLabs, Ltd. * All Rights Reserved. * */ @@ -329,12 +329,12 @@ struct hl_mem_in { struct { /* * Requested virtual address of mapped memory. - * KMD will try to map the requested region to this - * hint address, as long as the address is valid and - * not already mapped. The user should check the + * The driver will try to map the requested region to + * this hint address, as long as the address is valid + * and not already mapped. The user should check the * returned address of the IOCTL to make sure he got - * the hint address. Passing 0 here means that KMD - * will choose the address itself. + * the hint address. Passing 0 here means that the + * driver will choose the address itself. */ __u64 hint_addr; /* Handle returned from HL_MEM_OP_ALLOC */ @@ -347,12 +347,12 @@ struct hl_mem_in { __u64 host_virt_addr; /* * Requested virtual address of mapped memory. - * KMD will try to map the requested region to this - * hint address, as long as the address is valid and - * not already mapped. The user should check the + * The driver will try to map the requested region to + * this hint address, as long as the address is valid + * and not already mapped. The user should check the * returned address of the IOCTL to make sure he got - * the hint address. Passing 0 here means that KMD - * will choose the address itself. + * the hint address. Passing 0 here means that the + * driver will choose the address itself. */ __u64 hint_addr; /* Size of allocated host memory */ -- cgit v1.2.3-59-g8ed1b