From 7b4ccb3c466f62bbf2f4dd5d6a143d945a6f3051 Mon Sep 17 00:00:00 2001 From: Geert Uytterhoeven Date: Tue, 25 Apr 2017 19:36:25 +0200 Subject: soc: renesas: Provide dummy rcar_rst_read_mode_pins() for compile-testing If the R-Car RST driver is not included, compile-testing R-Car clock drivers fails with a link error: undefined reference to `rcar_rst_read_mode_pins' To fix this, provide a dummy version. Use the exact same test logic as in drivers/soc/renesas/Makefile, as there is no Kconfig symbol (yet) to control compilation of the R-Car RST driver. Fixes: 527c02f66d263d2e ("soc: renesas: Add R-Car RST driver") Signed-off-by: Geert Uytterhoeven Signed-off-by: Simon Horman --- include/linux/soc/renesas/rcar-rst.h | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'include') diff --git a/include/linux/soc/renesas/rcar-rst.h b/include/linux/soc/renesas/rcar-rst.h index a18e0783946b..787e7ad53d45 100644 --- a/include/linux/soc/renesas/rcar-rst.h +++ b/include/linux/soc/renesas/rcar-rst.h @@ -1,6 +1,11 @@ #ifndef __LINUX_SOC_RENESAS_RCAR_RST_H__ #define __LINUX_SOC_RENESAS_RCAR_RST_H__ +#if defined(CONFIG_ARCH_RCAR_GEN1) || defined(CONFIG_ARCH_RCAR_GEN2) || \ + defined(CONFIG_ARCH_R8A7795) || defined(CONFIG_ARCH_R8A7796) int rcar_rst_read_mode_pins(u32 *mode); +#else +static inline int rcar_rst_read_mode_pins(u32 *mode) { return -ENODEV; } +#endif #endif /* __LINUX_SOC_RENESAS_RCAR_RST_H__ */ -- cgit v1.3-14-g43fede From df3ed932394488e57e72dd0e73c224d1804fdc8f Mon Sep 17 00:00:00 2001 From: Rob Herring Date: Thu, 11 May 2017 10:15:10 -0500 Subject: Partially Revert "of: fix sparse warnings in fdt, irq, reserved mem, and resolver code" A change to function pointers that was meant to address a sparse warning turned out to cause hundreds of new gcc-7 warnings: include/linux/of_irq.h:11:13: error: type qualifiers ignored on function return type [-Werror=ignored-qualifiers] drivers/of/of_reserved_mem.c: In function '__reserved_mem_init_node': drivers/of/of_reserved_mem.c:200:7: error: type qualifiers ignored on function return type [-Werror=ignored-qualifiers] int const (*initfn)(struct reserved_mem *rmem) = i->data; Turns out the sparse warnings were spurious and have been fixed in upstream sparse since 0.5.0 in commit "sparse: treat function pointers as pointers to const data". This partially reverts commit 17a70355ea576843a7ac851f1db26872a50b2850. Fixes: 17a70355ea57 ("of: fix sparse warnings in fdt, irq, reserved mem, and resolver code") Reported-by: Arnd Bergmann Signed-off-by: Rob Herring --- drivers/of/of_reserved_mem.c | 2 +- include/linux/of_irq.h | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/drivers/of/of_reserved_mem.c b/drivers/of/of_reserved_mem.c index 4dec07ea510f..d507c3569a88 100644 --- a/drivers/of/of_reserved_mem.c +++ b/drivers/of/of_reserved_mem.c @@ -197,7 +197,7 @@ static int __init __reserved_mem_init_node(struct reserved_mem *rmem) const struct of_device_id *i; for (i = __reservedmem_of_table; i < &__rmem_of_table_sentinel; i++) { - int const (*initfn)(struct reserved_mem *rmem) = i->data; + reservedmem_of_init_fn initfn = i->data; const char *compat = i->compatible; if (!of_flat_dt_is_compatible(rmem->fdt_node, compat)) diff --git a/include/linux/of_irq.h b/include/linux/of_irq.h index ec6b11deb773..1e0deb8e8494 100644 --- a/include/linux/of_irq.h +++ b/include/linux/of_irq.h @@ -8,7 +8,7 @@ #include #include -typedef int const (*of_irq_init_cb_t)(struct device_node *, struct device_node *); +typedef int (*of_irq_init_cb_t)(struct device_node *, struct device_node *); /* * Workarounds only applied to 32bit powermac machines -- cgit v1.3-14-g43fede From f5705aa8cfed142d980ecac12bee0d81b756479e Mon Sep 17 00:00:00 2001 From: Dan Williams Date: Sat, 13 May 2017 16:31:05 -0700 Subject: dax, xfs, ext4: compile out iomap-dax paths in the FS_DAX=n case Tetsuo reports: fs/built-in.o: In function `xfs_file_iomap_end': xfs_iomap.c:(.text+0xe0ef9): undefined reference to `put_dax' fs/built-in.o: In function `xfs_file_iomap_begin': xfs_iomap.c:(.text+0xe1a7f): undefined reference to `dax_get_by_host' make: *** [vmlinux] Error 1 $ grep DAX .config CONFIG_DAX=m # CONFIG_DEV_DAX is not set # CONFIG_FS_DAX is not set When FS_DAX=n we can/must throw away the dax code in filesystems. Implement 'fs_' versions of dax_get_by_host() and put_dax() that are nops in the FS_DAX=n case. Cc: Cc: Cc: Jan Kara Cc: "Theodore Ts'o" Cc: "Darrick J. Wong" Cc: Ross Zwisler Tested-by: Tony Luck Fixes: ef51042472f5 ("block, dax: move 'select DAX' from BLOCK to FS_DAX") Reported-by: Tetsuo Handa Signed-off-by: Dan Williams --- fs/ext2/inode.c | 4 ++-- fs/ext4/inode.c | 4 ++-- fs/xfs/xfs_iomap.c | 4 ++-- include/linux/dax.h | 34 +++++++++++++++++++++++++++------- 4 files changed, 33 insertions(+), 13 deletions(-) (limited to 'include') diff --git a/fs/ext2/inode.c b/fs/ext2/inode.c index 26d77f9f8c12..2dcbd5698884 100644 --- a/fs/ext2/inode.c +++ b/fs/ext2/inode.c @@ -817,7 +817,7 @@ static int ext2_iomap_begin(struct inode *inode, loff_t offset, loff_t length, iomap->bdev = bdev; iomap->offset = (u64)first_block << blkbits; if (blk_queue_dax(bdev->bd_queue)) - iomap->dax_dev = dax_get_by_host(bdev->bd_disk->disk_name); + iomap->dax_dev = fs_dax_get_by_host(bdev->bd_disk->disk_name); else iomap->dax_dev = NULL; @@ -841,7 +841,7 @@ static int ext2_iomap_end(struct inode *inode, loff_t offset, loff_t length, ssize_t written, unsigned flags, struct iomap *iomap) { - put_dax(iomap->dax_dev); + fs_put_dax(iomap->dax_dev); if (iomap->type == IOMAP_MAPPED && written < length && (flags & IOMAP_WRITE)) diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 5834c4d76be8..1bd0bfa547f6 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -3412,7 +3412,7 @@ retry: bdev = inode->i_sb->s_bdev; iomap->bdev = bdev; if (blk_queue_dax(bdev->bd_queue)) - iomap->dax_dev = dax_get_by_host(bdev->bd_disk->disk_name); + iomap->dax_dev = fs_dax_get_by_host(bdev->bd_disk->disk_name); else iomap->dax_dev = NULL; iomap->offset = first_block << blkbits; @@ -3447,7 +3447,7 @@ static int ext4_iomap_end(struct inode *inode, loff_t offset, loff_t length, int blkbits = inode->i_blkbits; bool truncate = false; - put_dax(iomap->dax_dev); + fs_put_dax(iomap->dax_dev); if (!(flags & IOMAP_WRITE) || (flags & IOMAP_FAULT)) return 0; diff --git a/fs/xfs/xfs_iomap.c b/fs/xfs/xfs_iomap.c index a63f61c256bd..94e5bdf7304c 100644 --- a/fs/xfs/xfs_iomap.c +++ b/fs/xfs/xfs_iomap.c @@ -1068,7 +1068,7 @@ xfs_file_iomap_begin( /* optionally associate a dax device with the iomap bdev */ bdev = iomap->bdev; if (blk_queue_dax(bdev->bd_queue)) - iomap->dax_dev = dax_get_by_host(bdev->bd_disk->disk_name); + iomap->dax_dev = fs_dax_get_by_host(bdev->bd_disk->disk_name); else iomap->dax_dev = NULL; @@ -1149,7 +1149,7 @@ xfs_file_iomap_end( unsigned flags, struct iomap *iomap) { - put_dax(iomap->dax_dev); + fs_put_dax(iomap->dax_dev); if ((flags & IOMAP_WRITE) && iomap->type == IOMAP_DELALLOC) return xfs_file_iomap_end_delalloc(XFS_I(inode), offset, length, written, iomap); diff --git a/include/linux/dax.h b/include/linux/dax.h index 00ebac854bb7..5ec1f6c47716 100644 --- a/include/linux/dax.h +++ b/include/linux/dax.h @@ -18,6 +18,20 @@ struct dax_operations { void **, pfn_t *); }; +#if IS_ENABLED(CONFIG_DAX) +struct dax_device *dax_get_by_host(const char *host); +void put_dax(struct dax_device *dax_dev); +#else +static inline struct dax_device *dax_get_by_host(const char *host) +{ + return NULL; +} + +static inline void put_dax(struct dax_device *dax_dev) +{ +} +#endif + int bdev_dax_pgoff(struct block_device *, sector_t, size_t, pgoff_t *pgoff); #if IS_ENABLED(CONFIG_FS_DAX) int __bdev_dax_supported(struct super_block *sb, int blocksize); @@ -25,23 +39,29 @@ static inline int bdev_dax_supported(struct super_block *sb, int blocksize) { return __bdev_dax_supported(sb, blocksize); } + +static inline struct dax_device *fs_dax_get_by_host(const char *host) +{ + return dax_get_by_host(host); +} + +static inline void fs_put_dax(struct dax_device *dax_dev) +{ + put_dax(dax_dev); +} + #else static inline int bdev_dax_supported(struct super_block *sb, int blocksize) { return -EOPNOTSUPP; } -#endif -#if IS_ENABLED(CONFIG_DAX) -struct dax_device *dax_get_by_host(const char *host); -void put_dax(struct dax_device *dax_dev); -#else -static inline struct dax_device *dax_get_by_host(const char *host) +static inline struct dax_device *fs_dax_get_by_host(const char *host) { return NULL; } -static inline void put_dax(struct dax_device *dax_dev) +static inline void fs_put_dax(struct dax_device *dax_dev) { } #endif -- cgit v1.3-14-g43fede From d91fc59cd77c719f33eda65c194ad8f95a055190 Mon Sep 17 00:00:00 2001 From: Liping Zhang Date: Sun, 7 May 2017 22:01:55 +0800 Subject: netfilter: introduce nf_conntrack_helper_put helper function And convert module_put invocation to nf_conntrack_helper_put, this is prepared for the followup patch, which will add a refcnt for cthelper, so we can reject the deleting request when cthelper is in use. Signed-off-by: Liping Zhang Signed-off-by: Pablo Neira Ayuso --- include/net/netfilter/nf_conntrack_helper.h | 2 ++ net/netfilter/nf_conntrack_helper.c | 6 ++++++ net/netfilter/nft_ct.c | 4 ++-- net/netfilter/xt_CT.c | 6 +++--- net/openvswitch/conntrack.c | 4 ++-- 5 files changed, 15 insertions(+), 7 deletions(-) (limited to 'include') diff --git a/include/net/netfilter/nf_conntrack_helper.h b/include/net/netfilter/nf_conntrack_helper.h index e04fa7691e5d..c1c12411103a 100644 --- a/include/net/netfilter/nf_conntrack_helper.h +++ b/include/net/netfilter/nf_conntrack_helper.h @@ -79,6 +79,8 @@ struct nf_conntrack_helper *__nf_conntrack_helper_find(const char *name, struct nf_conntrack_helper *nf_conntrack_helper_try_module_get(const char *name, u16 l3num, u8 protonum); +void nf_conntrack_helper_put(struct nf_conntrack_helper *helper); + void nf_ct_helper_init(struct nf_conntrack_helper *helper, u16 l3num, u16 protonum, const char *name, u16 default_port, u16 spec_port, u32 id, diff --git a/net/netfilter/nf_conntrack_helper.c b/net/netfilter/nf_conntrack_helper.c index 3a60efa7799b..e17006b6e434 100644 --- a/net/netfilter/nf_conntrack_helper.c +++ b/net/netfilter/nf_conntrack_helper.c @@ -181,6 +181,12 @@ nf_conntrack_helper_try_module_get(const char *name, u16 l3num, u8 protonum) } EXPORT_SYMBOL_GPL(nf_conntrack_helper_try_module_get); +void nf_conntrack_helper_put(struct nf_conntrack_helper *helper) +{ + module_put(helper->me); +} +EXPORT_SYMBOL_GPL(nf_conntrack_helper_put); + struct nf_conn_help * nf_ct_helper_ext_add(struct nf_conn *ct, struct nf_conntrack_helper *helper, gfp_t gfp) diff --git a/net/netfilter/nft_ct.c b/net/netfilter/nft_ct.c index a34ceb38fc55..1678e9e75e8e 100644 --- a/net/netfilter/nft_ct.c +++ b/net/netfilter/nft_ct.c @@ -826,9 +826,9 @@ static void nft_ct_helper_obj_destroy(struct nft_object *obj) struct nft_ct_helper_obj *priv = nft_obj_data(obj); if (priv->helper4) - module_put(priv->helper4->me); + nf_conntrack_helper_put(priv->helper4); if (priv->helper6) - module_put(priv->helper6->me); + nf_conntrack_helper_put(priv->helper6); } static void nft_ct_helper_obj_eval(struct nft_object *obj, diff --git a/net/netfilter/xt_CT.c b/net/netfilter/xt_CT.c index bb7ad82dcd56..623ef37de886 100644 --- a/net/netfilter/xt_CT.c +++ b/net/netfilter/xt_CT.c @@ -96,7 +96,7 @@ xt_ct_set_helper(struct nf_conn *ct, const char *helper_name, help = nf_ct_helper_ext_add(ct, helper, GFP_KERNEL); if (help == NULL) { - module_put(helper->me); + nf_conntrack_helper_put(helper); return -ENOMEM; } @@ -263,7 +263,7 @@ out: err4: help = nfct_help(ct); if (help) - module_put(help->helper->me); + nf_conntrack_helper_put(help->helper); err3: nf_ct_tmpl_free(ct); err2: @@ -346,7 +346,7 @@ static void xt_ct_tg_destroy(const struct xt_tgdtor_param *par, if (ct) { help = nfct_help(ct); if (help) - module_put(help->helper->me); + nf_conntrack_helper_put(help->helper); nf_ct_netns_put(par->net, par->family); diff --git a/net/openvswitch/conntrack.c b/net/openvswitch/conntrack.c index bf602e33c40a..08679ebb3068 100644 --- a/net/openvswitch/conntrack.c +++ b/net/openvswitch/conntrack.c @@ -1123,7 +1123,7 @@ static int ovs_ct_add_helper(struct ovs_conntrack_info *info, const char *name, help = nf_ct_helper_ext_add(info->ct, helper, GFP_KERNEL); if (!help) { - module_put(helper->me); + nf_conntrack_helper_put(helper); return -ENOMEM; } @@ -1584,7 +1584,7 @@ void ovs_ct_free_action(const struct nlattr *a) static void __ovs_ct_free_action(struct ovs_conntrack_info *ct_info) { if (ct_info->helper) - module_put(ct_info->helper->me); + nf_conntrack_helper_put(ct_info->helper); if (ct_info->ct) nf_ct_tmpl_free(ct_info->ct); } -- cgit v1.3-14-g43fede From 9338d7b4418e9996a7642867d8f6b482a6040ed6 Mon Sep 17 00:00:00 2001 From: Liping Zhang Date: Sun, 7 May 2017 22:01:56 +0800 Subject: netfilter: nfnl_cthelper: reject del request if helper obj is in use We can still delete the ct helper even if it is in use, this will cause a use-after-free error. In more detail, I mean: # nfct helper add ssdp inet udp # iptables -t raw -A OUTPUT -p udp -j CT --helper ssdp # nfct helper delete ssdp //--> oops, succeed! BUG: unable to handle kernel paging request at 000026ca IP: 0x26ca [...] Call Trace: ? ipv4_helper+0x62/0x80 [nf_conntrack_ipv4] nf_hook_slow+0x21/0xb0 ip_output+0xe9/0x100 ? ip_fragment.constprop.54+0xc0/0xc0 ip_local_out+0x33/0x40 ip_send_skb+0x16/0x80 udp_send_skb+0x84/0x240 udp_sendmsg+0x35d/0xa50 So add reference count to fix this issue, if ct helper is used by others, reject the delete request. Apply this patch: # nfct helper delete ssdp nfct v1.4.3: netlink error: Device or resource busy Signed-off-by: Liping Zhang Signed-off-by: Pablo Neira Ayuso --- include/net/netfilter/nf_conntrack_helper.h | 2 ++ net/netfilter/nf_conntrack_helper.c | 6 ++++++ net/netfilter/nfnetlink_cthelper.c | 17 +++++++++++------ 3 files changed, 19 insertions(+), 6 deletions(-) (limited to 'include') diff --git a/include/net/netfilter/nf_conntrack_helper.h b/include/net/netfilter/nf_conntrack_helper.h index c1c12411103a..c519bb5b5bb8 100644 --- a/include/net/netfilter/nf_conntrack_helper.h +++ b/include/net/netfilter/nf_conntrack_helper.h @@ -9,6 +9,7 @@ #ifndef _NF_CONNTRACK_HELPER_H #define _NF_CONNTRACK_HELPER_H +#include #include #include #include @@ -26,6 +27,7 @@ struct nf_conntrack_helper { struct hlist_node hnode; /* Internal use. */ char name[NF_CT_HELPER_NAME_LEN]; /* name of the module */ + refcount_t refcnt; struct module *me; /* pointer to self */ const struct nf_conntrack_expect_policy *expect_policy; diff --git a/net/netfilter/nf_conntrack_helper.c b/net/netfilter/nf_conntrack_helper.c index e17006b6e434..7f6100ca63be 100644 --- a/net/netfilter/nf_conntrack_helper.c +++ b/net/netfilter/nf_conntrack_helper.c @@ -174,6 +174,10 @@ nf_conntrack_helper_try_module_get(const char *name, u16 l3num, u8 protonum) #endif if (h != NULL && !try_module_get(h->me)) h = NULL; + if (h != NULL && !refcount_inc_not_zero(&h->refcnt)) { + module_put(h->me); + h = NULL; + } rcu_read_unlock(); @@ -183,6 +187,7 @@ EXPORT_SYMBOL_GPL(nf_conntrack_helper_try_module_get); void nf_conntrack_helper_put(struct nf_conntrack_helper *helper) { + refcount_dec(&helper->refcnt); module_put(helper->me); } EXPORT_SYMBOL_GPL(nf_conntrack_helper_put); @@ -423,6 +428,7 @@ int nf_conntrack_helper_register(struct nf_conntrack_helper *me) } } } + refcount_set(&me->refcnt, 1); hlist_add_head_rcu(&me->hnode, &nf_ct_helper_hash[h]); nf_ct_helper_count++; out: diff --git a/net/netfilter/nfnetlink_cthelper.c b/net/netfilter/nfnetlink_cthelper.c index 950bf6eadc65..be678a323598 100644 --- a/net/netfilter/nfnetlink_cthelper.c +++ b/net/netfilter/nfnetlink_cthelper.c @@ -686,6 +686,7 @@ static int nfnl_cthelper_del(struct net *net, struct sock *nfnl, tuple_set = true; } + ret = -ENOENT; list_for_each_entry_safe(nlcth, n, &nfnl_cthelper_list, list) { cur = &nlcth->helper; j++; @@ -699,16 +700,20 @@ static int nfnl_cthelper_del(struct net *net, struct sock *nfnl, tuple.dst.protonum != cur->tuple.dst.protonum)) continue; - found = true; - nf_conntrack_helper_unregister(cur); - kfree(cur->expect_policy); + if (refcount_dec_if_one(&cur->refcnt)) { + found = true; + nf_conntrack_helper_unregister(cur); + kfree(cur->expect_policy); - list_del(&nlcth->list); - kfree(nlcth); + list_del(&nlcth->list); + kfree(nlcth); + } else { + ret = -EBUSY; + } } /* Make sure we return success if we flush and there is no helpers */ - return (found || j == 0) ? 0 : -ENOENT; + return (found || j == 0) ? 0 : ret; } static const struct nla_policy nfnl_cthelper_policy[NFCTH_MAX+1] = { -- cgit v1.3-14-g43fede From 324318f0248c31be8a08984146e7e4dd7cdd091d Mon Sep 17 00:00:00 2001 From: Willem de Bruijn Date: Tue, 9 May 2017 16:17:37 -0400 Subject: netfilter: xtables: zero padding in data_to_user When looking up an iptables rule, the iptables binary compares the aligned match and target data (XT_ALIGN). In some cases this can exceed the actual data size to include padding bytes. Before commit f77bc5b23fb1 ("iptables: use match, target and data copy_to_user helpers") the malloc()ed bytes were overwritten by the kernel with kzalloced contents, zeroing the padding and making the comparison succeed. After this patch, the kernel copies and clears only data, leaving the padding bytes undefined. Extend the clear operation from data size to aligned data size to include the padding bytes, if any. Padding bytes can be observed in both match and target, and the bug triggered, by issuing a rule with match icmp and target ACCEPT: iptables -t mangle -A INPUT -i lo -p icmp --icmp-type 1 -j ACCEPT iptables -t mangle -D INPUT -i lo -p icmp --icmp-type 1 -j ACCEPT Fixes: f77bc5b23fb1 ("iptables: use match, target and data copy_to_user helpers") Reported-by: Paul Moore Reported-by: Richard Guy Briggs Signed-off-by: Willem de Bruijn Signed-off-by: Pablo Neira Ayuso --- include/linux/netfilter/x_tables.h | 2 +- net/bridge/netfilter/ebtables.c | 9 ++++++--- net/netfilter/x_tables.c | 9 ++++++--- 3 files changed, 13 insertions(+), 7 deletions(-) (limited to 'include') diff --git a/include/linux/netfilter/x_tables.h b/include/linux/netfilter/x_tables.h index be378cf47fcc..b3044c2c62cb 100644 --- a/include/linux/netfilter/x_tables.h +++ b/include/linux/netfilter/x_tables.h @@ -294,7 +294,7 @@ int xt_match_to_user(const struct xt_entry_match *m, int xt_target_to_user(const struct xt_entry_target *t, struct xt_entry_target __user *u); int xt_data_to_user(void __user *dst, const void *src, - int usersize, int size); + int usersize, int size, int aligned_size); void *xt_copy_counters_from_user(const void __user *user, unsigned int len, struct xt_counters_info *info, bool compat); diff --git a/net/bridge/netfilter/ebtables.c b/net/bridge/netfilter/ebtables.c index 9ec0c9f908fa..9c6e619f452b 100644 --- a/net/bridge/netfilter/ebtables.c +++ b/net/bridge/netfilter/ebtables.c @@ -1373,7 +1373,8 @@ static inline int ebt_obj_to_user(char __user *um, const char *_name, strlcpy(name, _name, sizeof(name)); if (copy_to_user(um, name, EBT_FUNCTION_MAXNAMELEN) || put_user(datasize, (int __user *)(um + EBT_FUNCTION_MAXNAMELEN)) || - xt_data_to_user(um + entrysize, data, usersize, datasize)) + xt_data_to_user(um + entrysize, data, usersize, datasize, + XT_ALIGN(datasize))) return -EFAULT; return 0; @@ -1658,7 +1659,8 @@ static int compat_match_to_user(struct ebt_entry_match *m, void __user **dstptr, if (match->compat_to_user(cm->data, m->data)) return -EFAULT; } else { - if (xt_data_to_user(cm->data, m->data, match->usersize, msize)) + if (xt_data_to_user(cm->data, m->data, match->usersize, msize, + COMPAT_XT_ALIGN(msize))) return -EFAULT; } @@ -1687,7 +1689,8 @@ static int compat_target_to_user(struct ebt_entry_target *t, if (target->compat_to_user(cm->data, t->data)) return -EFAULT; } else { - if (xt_data_to_user(cm->data, t->data, target->usersize, tsize)) + if (xt_data_to_user(cm->data, t->data, target->usersize, tsize, + COMPAT_XT_ALIGN(tsize))) return -EFAULT; } diff --git a/net/netfilter/x_tables.c b/net/netfilter/x_tables.c index 8876b7da6884..d17769599c10 100644 --- a/net/netfilter/x_tables.c +++ b/net/netfilter/x_tables.c @@ -283,12 +283,13 @@ static int xt_obj_to_user(u16 __user *psize, u16 size, &U->u.user.revision, K->u.kernel.TYPE->revision) int xt_data_to_user(void __user *dst, const void *src, - int usersize, int size) + int usersize, int size, int aligned_size) { usersize = usersize ? : size; if (copy_to_user(dst, src, usersize)) return -EFAULT; - if (usersize != size && clear_user(dst + usersize, size - usersize)) + if (usersize != aligned_size && + clear_user(dst + usersize, aligned_size - usersize)) return -EFAULT; return 0; @@ -298,7 +299,9 @@ EXPORT_SYMBOL_GPL(xt_data_to_user); #define XT_DATA_TO_USER(U, K, TYPE, C_SIZE) \ xt_data_to_user(U->data, K->data, \ K->u.kernel.TYPE->usersize, \ - C_SIZE ? : K->u.kernel.TYPE->TYPE##size) + C_SIZE ? : K->u.kernel.TYPE->TYPE##size, \ + C_SIZE ? COMPAT_XT_ALIGN(C_SIZE) : \ + XT_ALIGN(K->u.kernel.TYPE->TYPE##size)) int xt_match_to_user(const struct xt_entry_match *m, struct xt_entry_match __user *u) -- cgit v1.3-14-g43fede From 591054469b3eef34bc097c30fae8ededddf8d796 Mon Sep 17 00:00:00 2001 From: Pablo Neira Ayuso Date: Mon, 15 May 2017 11:17:34 +0100 Subject: netfilter: nf_tables: revisit chain/object refcounting from elements Andreas reports that the following incremental update using our commit protocol doesn't work. # nft -f incremental-update.nft delete element ip filter client_to_any { 10.180.86.22 : goto CIn_1 } delete chain ip filter CIn_1 ... Error: Could not process rule: Device or resource busy The existing code is not well-integrated into the commit phase protocol, since element deletions do not result in refcount decrement from the preparation phase. This results in bogus EBUSY errors like the one above. Two new functions come with this patch: * nft_set_elem_activate() function is used from the abort path, to restore the set element refcounting on objects that occurred from the preparation phase. * nft_set_elem_deactivate() that is called from nft_del_setelem() to decrement set element refcounting on objects from the preparation phase in the commit protocol. The nft_data_uninit() has been renamed to nft_data_release() since this function does not uninitialize any data store in the data register, instead just releases the references to objects. Moreover, a new function nft_data_hold() has been introduced to be used from nft_set_elem_activate(). Reported-by: Andreas Schultz Signed-off-by: Pablo Neira Ayuso --- include/net/netfilter/nf_tables.h | 2 +- net/netfilter/nf_tables_api.c | 82 ++++++++++++++++++++++++++++++++++----- net/netfilter/nft_bitwise.c | 4 +- net/netfilter/nft_cmp.c | 2 +- net/netfilter/nft_immediate.c | 5 ++- net/netfilter/nft_range.c | 4 +- 6 files changed, 81 insertions(+), 18 deletions(-) (limited to 'include') diff --git a/include/net/netfilter/nf_tables.h b/include/net/netfilter/nf_tables.h index 028faec8fc27..8a8bab8d7b15 100644 --- a/include/net/netfilter/nf_tables.h +++ b/include/net/netfilter/nf_tables.h @@ -176,7 +176,7 @@ struct nft_data_desc { int nft_data_init(const struct nft_ctx *ctx, struct nft_data *data, unsigned int size, struct nft_data_desc *desc, const struct nlattr *nla); -void nft_data_uninit(const struct nft_data *data, enum nft_data_types type); +void nft_data_release(const struct nft_data *data, enum nft_data_types type); int nft_data_dump(struct sk_buff *skb, int attr, const struct nft_data *data, enum nft_data_types type, unsigned int len); diff --git a/net/netfilter/nf_tables_api.c b/net/netfilter/nf_tables_api.c index 5f4a4d48b871..da314be0c048 100644 --- a/net/netfilter/nf_tables_api.c +++ b/net/netfilter/nf_tables_api.c @@ -3627,9 +3627,9 @@ void nft_set_elem_destroy(const struct nft_set *set, void *elem, { struct nft_set_ext *ext = nft_set_elem_ext(set, elem); - nft_data_uninit(nft_set_ext_key(ext), NFT_DATA_VALUE); + nft_data_release(nft_set_ext_key(ext), NFT_DATA_VALUE); if (nft_set_ext_exists(ext, NFT_SET_EXT_DATA)) - nft_data_uninit(nft_set_ext_data(ext), set->dtype); + nft_data_release(nft_set_ext_data(ext), set->dtype); if (destroy_expr && nft_set_ext_exists(ext, NFT_SET_EXT_EXPR)) nf_tables_expr_destroy(NULL, nft_set_ext_expr(ext)); if (nft_set_ext_exists(ext, NFT_SET_EXT_OBJREF)) @@ -3638,6 +3638,18 @@ void nft_set_elem_destroy(const struct nft_set *set, void *elem, } EXPORT_SYMBOL_GPL(nft_set_elem_destroy); +/* Only called from commit path, nft_set_elem_deactivate() already deals with + * the refcounting from the preparation phase. + */ +static void nf_tables_set_elem_destroy(const struct nft_set *set, void *elem) +{ + struct nft_set_ext *ext = nft_set_elem_ext(set, elem); + + if (nft_set_ext_exists(ext, NFT_SET_EXT_EXPR)) + nf_tables_expr_destroy(NULL, nft_set_ext_expr(ext)); + kfree(elem); +} + static int nft_setelem_parse_flags(const struct nft_set *set, const struct nlattr *attr, u32 *flags) { @@ -3849,9 +3861,9 @@ err4: kfree(elem.priv); err3: if (nla[NFTA_SET_ELEM_DATA] != NULL) - nft_data_uninit(&data, d2.type); + nft_data_release(&data, d2.type); err2: - nft_data_uninit(&elem.key.val, d1.type); + nft_data_release(&elem.key.val, d1.type); err1: return err; } @@ -3896,6 +3908,53 @@ static int nf_tables_newsetelem(struct net *net, struct sock *nlsk, return err; } +/** + * nft_data_hold - hold a nft_data item + * + * @data: struct nft_data to release + * @type: type of data + * + * Hold a nft_data item. NFT_DATA_VALUE types can be silently discarded, + * NFT_DATA_VERDICT bumps the reference to chains in case of NFT_JUMP and + * NFT_GOTO verdicts. This function must be called on active data objects + * from the second phase of the commit protocol. + */ +static void nft_data_hold(const struct nft_data *data, enum nft_data_types type) +{ + if (type == NFT_DATA_VERDICT) { + switch (data->verdict.code) { + case NFT_JUMP: + case NFT_GOTO: + data->verdict.chain->use++; + break; + } + } +} + +static void nft_set_elem_activate(const struct net *net, + const struct nft_set *set, + struct nft_set_elem *elem) +{ + const struct nft_set_ext *ext = nft_set_elem_ext(set, elem->priv); + + if (nft_set_ext_exists(ext, NFT_SET_EXT_DATA)) + nft_data_hold(nft_set_ext_data(ext), set->dtype); + if (nft_set_ext_exists(ext, NFT_SET_EXT_OBJREF)) + (*nft_set_ext_obj(ext))->use++; +} + +static void nft_set_elem_deactivate(const struct net *net, + const struct nft_set *set, + struct nft_set_elem *elem) +{ + const struct nft_set_ext *ext = nft_set_elem_ext(set, elem->priv); + + if (nft_set_ext_exists(ext, NFT_SET_EXT_DATA)) + nft_data_release(nft_set_ext_data(ext), set->dtype); + if (nft_set_ext_exists(ext, NFT_SET_EXT_OBJREF)) + (*nft_set_ext_obj(ext))->use--; +} + static int nft_del_setelem(struct nft_ctx *ctx, struct nft_set *set, const struct nlattr *attr) { @@ -3961,6 +4020,8 @@ static int nft_del_setelem(struct nft_ctx *ctx, struct nft_set *set, kfree(elem.priv); elem.priv = priv; + nft_set_elem_deactivate(ctx->net, set, &elem); + nft_trans_elem(trans) = elem; list_add_tail(&trans->list, &ctx->net->nft.commit_list); return 0; @@ -3970,7 +4031,7 @@ err4: err3: kfree(elem.priv); err2: - nft_data_uninit(&elem.key.val, desc.type); + nft_data_release(&elem.key.val, desc.type); err1: return err; } @@ -4777,8 +4838,8 @@ static void nf_tables_commit_release(struct nft_trans *trans) nft_set_destroy(nft_trans_set(trans)); break; case NFT_MSG_DELSETELEM: - nft_set_elem_destroy(nft_trans_elem_set(trans), - nft_trans_elem(trans).priv, true); + nf_tables_set_elem_destroy(nft_trans_elem_set(trans), + nft_trans_elem(trans).priv); break; case NFT_MSG_DELOBJ: nft_obj_destroy(nft_trans_obj(trans)); @@ -5013,6 +5074,7 @@ static int nf_tables_abort(struct net *net, struct sk_buff *skb) case NFT_MSG_DELSETELEM: te = (struct nft_trans_elem *)trans->data; + nft_set_elem_activate(net, te->set, &te->elem); te->set->ops->activate(net, te->set, &te->elem); te->set->ndeact--; @@ -5498,7 +5560,7 @@ int nft_data_init(const struct nft_ctx *ctx, EXPORT_SYMBOL_GPL(nft_data_init); /** - * nft_data_uninit - release a nft_data item + * nft_data_release - release a nft_data item * * @data: struct nft_data to release * @type: type of data @@ -5506,7 +5568,7 @@ EXPORT_SYMBOL_GPL(nft_data_init); * Release a nft_data item. NFT_DATA_VALUE types can be silently discarded, * all others need to be released by calling this function. */ -void nft_data_uninit(const struct nft_data *data, enum nft_data_types type) +void nft_data_release(const struct nft_data *data, enum nft_data_types type) { if (type < NFT_DATA_VERDICT) return; @@ -5517,7 +5579,7 @@ void nft_data_uninit(const struct nft_data *data, enum nft_data_types type) WARN_ON(1); } } -EXPORT_SYMBOL_GPL(nft_data_uninit); +EXPORT_SYMBOL_GPL(nft_data_release); int nft_data_dump(struct sk_buff *skb, int attr, const struct nft_data *data, enum nft_data_types type, unsigned int len) diff --git a/net/netfilter/nft_bitwise.c b/net/netfilter/nft_bitwise.c index 96bd4f325b0f..fff8073e2a56 100644 --- a/net/netfilter/nft_bitwise.c +++ b/net/netfilter/nft_bitwise.c @@ -99,9 +99,9 @@ static int nft_bitwise_init(const struct nft_ctx *ctx, return 0; err2: - nft_data_uninit(&priv->xor, d2.type); + nft_data_release(&priv->xor, d2.type); err1: - nft_data_uninit(&priv->mask, d1.type); + nft_data_release(&priv->mask, d1.type); return err; } diff --git a/net/netfilter/nft_cmp.c b/net/netfilter/nft_cmp.c index 8c9d0fb19118..c2945eb3397c 100644 --- a/net/netfilter/nft_cmp.c +++ b/net/netfilter/nft_cmp.c @@ -211,7 +211,7 @@ nft_cmp_select_ops(const struct nft_ctx *ctx, const struct nlattr * const tb[]) return &nft_cmp_ops; err1: - nft_data_uninit(&data, desc.type); + nft_data_release(&data, desc.type); return ERR_PTR(-EINVAL); } diff --git a/net/netfilter/nft_immediate.c b/net/netfilter/nft_immediate.c index 728baf88295a..4717d7796927 100644 --- a/net/netfilter/nft_immediate.c +++ b/net/netfilter/nft_immediate.c @@ -65,7 +65,7 @@ static int nft_immediate_init(const struct nft_ctx *ctx, return 0; err1: - nft_data_uninit(&priv->data, desc.type); + nft_data_release(&priv->data, desc.type); return err; } @@ -73,7 +73,8 @@ static void nft_immediate_destroy(const struct nft_ctx *ctx, const struct nft_expr *expr) { const struct nft_immediate_expr *priv = nft_expr_priv(expr); - return nft_data_uninit(&priv->data, nft_dreg_to_type(priv->dreg)); + + return nft_data_release(&priv->data, nft_dreg_to_type(priv->dreg)); } static int nft_immediate_dump(struct sk_buff *skb, const struct nft_expr *expr) diff --git a/net/netfilter/nft_range.c b/net/netfilter/nft_range.c index 9edc74eedc10..cedb96c3619f 100644 --- a/net/netfilter/nft_range.c +++ b/net/netfilter/nft_range.c @@ -102,9 +102,9 @@ static int nft_range_init(const struct nft_ctx *ctx, const struct nft_expr *expr priv->len = desc_from.len; return 0; err2: - nft_data_uninit(&priv->data_to, desc_to.type); + nft_data_release(&priv->data_to, desc_to.type); err1: - nft_data_uninit(&priv->data_from, desc_from.type); + nft_data_release(&priv->data_from, desc_from.type); return err; } -- cgit v1.3-14-g43fede From c953d63548207a085abcb12a15fefc8a11ffdf0a Mon Sep 17 00:00:00 2001 From: Gao Feng Date: Tue, 16 May 2017 09:30:18 +0800 Subject: ebtables: arpreply: Add the standard target sanity check The info->target comes from userspace and it would be used directly. So we need to add the sanity check to make sure it is a valid standard target, although the ebtables tool has already checked it. Kernel needs to validate anything coming from userspace. If the target is set as an evil value, it would break the ebtables and cause a panic. Because the non-standard target is treated as one offset. Now add one helper function ebt_invalid_target, and we would replace the macro INVALID_TARGET later. Signed-off-by: Gao Feng Signed-off-by: Pablo Neira Ayuso --- include/linux/netfilter_bridge/ebtables.h | 5 +++++ net/bridge/netfilter/ebt_arpreply.c | 3 +++ 2 files changed, 8 insertions(+) (limited to 'include') diff --git a/include/linux/netfilter_bridge/ebtables.h b/include/linux/netfilter_bridge/ebtables.h index a30efb437e6d..e0cbf17af780 100644 --- a/include/linux/netfilter_bridge/ebtables.h +++ b/include/linux/netfilter_bridge/ebtables.h @@ -125,4 +125,9 @@ extern unsigned int ebt_do_table(struct sk_buff *skb, /* True if the target is not a standard target */ #define INVALID_TARGET (info->target < -NUM_STANDARD_TARGETS || info->target >= 0) +static inline bool ebt_invalid_target(int target) +{ + return (target < -NUM_STANDARD_TARGETS || target >= 0); +} + #endif diff --git a/net/bridge/netfilter/ebt_arpreply.c b/net/bridge/netfilter/ebt_arpreply.c index 5929309beaa1..db85230e49c3 100644 --- a/net/bridge/netfilter/ebt_arpreply.c +++ b/net/bridge/netfilter/ebt_arpreply.c @@ -68,6 +68,9 @@ static int ebt_arpreply_tg_check(const struct xt_tgchk_param *par) if (e->ethproto != htons(ETH_P_ARP) || e->invflags & EBT_IPROTO) return -EINVAL; + if (ebt_invalid_target(info->target)) + return -EINVAL; + return 0; } -- cgit v1.3-14-g43fede From 93491ced3c87c94b12220dbac0527e1356702179 Mon Sep 17 00:00:00 2001 From: Johan Hovold Date: Wed, 10 May 2017 18:18:29 +0200 Subject: USB: hub: fix SS max number of ports Add define for the maximum number of ports on a SuperSpeed hub as per USB 3.1 spec Table 10-5, and use it when verifying the retrieved hub descriptor. This specifically avoids benign attempts to update the DeviceRemovable mask for non-existing ports (should we get that far). Fixes: dbe79bbe9dcb ("USB 3.0 Hub Changes") Acked-by: Alan Stern Signed-off-by: Johan Hovold Signed-off-by: Greg Kroah-Hartman --- drivers/usb/core/hub.c | 8 +++++++- include/uapi/linux/usb/ch11.h | 3 +++ 2 files changed, 10 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/drivers/usb/core/hub.c b/drivers/usb/core/hub.c index f77a4ebde7d5..b8bb20d7acdb 100644 --- a/drivers/usb/core/hub.c +++ b/drivers/usb/core/hub.c @@ -1336,7 +1336,13 @@ static int hub_configure(struct usb_hub *hub, if (ret < 0) { message = "can't read hub descriptor"; goto fail; - } else if (hub->descriptor->bNbrPorts > USB_MAXCHILDREN) { + } + + maxchild = USB_MAXCHILDREN; + if (hub_is_superspeed(hdev)) + maxchild = min_t(unsigned, maxchild, USB_SS_MAXPORTS); + + if (hub->descriptor->bNbrPorts > maxchild) { message = "hub has too many ports!"; ret = -ENODEV; goto fail; diff --git a/include/uapi/linux/usb/ch11.h b/include/uapi/linux/usb/ch11.h index 361297e96f58..576c704e3fb8 100644 --- a/include/uapi/linux/usb/ch11.h +++ b/include/uapi/linux/usb/ch11.h @@ -22,6 +22,9 @@ */ #define USB_MAXCHILDREN 31 +/* See USB 3.1 spec Table 10-5 */ +#define USB_SS_MAXPORTS 15 + /* * Hub request types */ -- cgit v1.3-14-g43fede From 30e7d894c1478c88d50ce94ddcdbd7f9763d9cdd Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Wed, 17 May 2017 10:19:49 +0200 Subject: tracing/kprobes: Enforce kprobes teardown after testing Enabling the tracer selftest triggers occasionally the warning in text_poke(), which warns when the to be modified page is not marked reserved. The reason is that the tracer selftest installs kprobes on functions marked __init for testing. These probes are removed after the tests, but that removal schedules the delayed kprobes_optimizer work, which will do the actual text poke. If the work is executed after the init text is freed, then the warning triggers. The bug can be reproduced reliably when the work delay is increased. Flush the optimizer work and wait for the optimizing/unoptimizing lists to become empty before returning from the kprobes tracer selftest. That ensures that all operations which were queued due to the probes removal have completed. Link: http://lkml.kernel.org/r/20170516094802.76a468bb@gandalf.local.home Signed-off-by: Thomas Gleixner Acked-by: Masami Hiramatsu Cc: stable@vger.kernel.org Fixes: 6274de498 ("kprobes: Support delayed unoptimizing") Signed-off-by: Steven Rostedt (VMware) --- include/linux/kprobes.h | 3 +++ kernel/kprobes.c | 2 +- kernel/trace/trace_kprobe.c | 5 +++++ 3 files changed, 9 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/kprobes.h b/include/linux/kprobes.h index 30f90c1a0aaf..541df0b5b815 100644 --- a/include/linux/kprobes.h +++ b/include/linux/kprobes.h @@ -349,6 +349,9 @@ extern int proc_kprobes_optimization_handler(struct ctl_table *table, int write, void __user *buffer, size_t *length, loff_t *ppos); #endif +extern void wait_for_kprobe_optimizer(void); +#else +static inline void wait_for_kprobe_optimizer(void) { } #endif /* CONFIG_OPTPROBES */ #ifdef CONFIG_KPROBES_ON_FTRACE extern void kprobe_ftrace_handler(unsigned long ip, unsigned long parent_ip, diff --git a/kernel/kprobes.c b/kernel/kprobes.c index 7367e0ec6f81..199243bba554 100644 --- a/kernel/kprobes.c +++ b/kernel/kprobes.c @@ -595,7 +595,7 @@ static void kprobe_optimizer(struct work_struct *work) } /* Wait for completing optimization and unoptimization */ -static void wait_for_kprobe_optimizer(void) +void wait_for_kprobe_optimizer(void) { mutex_lock(&kprobe_mutex); diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index 8485f6738a87..c129fca6ec99 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -1535,6 +1535,11 @@ static __init int kprobe_trace_self_tests_init(void) end: release_all_trace_kprobes(); + /* + * Wait for the optimizer work to finish. Otherwise it might fiddle + * with probes in already freed __init text. + */ + wait_for_kprobe_optimizer(); if (warn) pr_cont("NG: Some tests are failed. Please check them.\n"); else -- cgit v1.3-14-g43fede From 552c9f47f8d451830a6b47151c6d2db77f77cc3e Mon Sep 17 00:00:00 2001 From: Christoffer Dall Date: Wed, 17 May 2017 13:12:51 +0200 Subject: KVM: arm/arm64: Fix bug when registering redist iodevs If userspace creates the VCPUs after initializing the VGIC, then we end up in a situation where we trigger a bug in kvm_vcpu_get_idx(), because it is called prior to adding the VCPU into the vcpus array on the VM. There is no tight coupling between the VCPU index and the area of the redistributor region used for the VCPU, so we can simply ensure that all creations of redistributors are serialized per VM, and increment an offset when we successfully add a redistributor. The vgic_register_redist_iodev() function can be called from two paths: vgic_redister_all_redist_iodev() which is called via the kvm_vgic_addr() device attribute handler. This patch already holds the kvm->lock mutex. The other path is via kvm_vgic_vcpu_init, which is called through a longer chain from kvm_vm_ioctl_create_vcpu(), which releases the kvm->lock mutex just before calling kvm_arch_vcpu_create(), so we can simply take this mutex again later for our purposes. Fixes: ab6f468c10 ("KVM: arm/arm64: Register iodevs when setting redist base and creating VCPUs") Signed-off-by: Christoffer Dall Tested-by: Jean-Philippe Brucker Reviewed-by: Eric Auger --- include/kvm/arm_vgic.h | 5 ++++- virt/kvm/arm/vgic/vgic-init.c | 5 ++++- virt/kvm/arm/vgic/vgic-mmio-v3.c | 9 ++++++--- 3 files changed, 14 insertions(+), 5 deletions(-) (limited to 'include') diff --git a/include/kvm/arm_vgic.h b/include/kvm/arm_vgic.h index 97b8d3728b31..ef718586321c 100644 --- a/include/kvm/arm_vgic.h +++ b/include/kvm/arm_vgic.h @@ -195,7 +195,10 @@ struct vgic_dist { /* either a GICv2 CPU interface */ gpa_t vgic_cpu_base; /* or a number of GICv3 redistributor regions */ - gpa_t vgic_redist_base; + struct { + gpa_t vgic_redist_base; + gpa_t vgic_redist_free_offset; + }; }; /* distributor enabled */ diff --git a/virt/kvm/arm/vgic/vgic-init.c b/virt/kvm/arm/vgic/vgic-init.c index dc68e2e424ab..3a0b8999f011 100644 --- a/virt/kvm/arm/vgic/vgic-init.c +++ b/virt/kvm/arm/vgic/vgic-init.c @@ -242,8 +242,11 @@ int kvm_vgic_vcpu_init(struct kvm_vcpu *vcpu) * If we are creating a VCPU with a GICv3 we must also register the * KVM io device for the redistributor that belongs to this VCPU. */ - if (dist->vgic_model == KVM_DEV_TYPE_ARM_VGIC_V3) + if (dist->vgic_model == KVM_DEV_TYPE_ARM_VGIC_V3) { + mutex_lock(&vcpu->kvm->lock); ret = vgic_register_redist_iodev(vcpu); + mutex_unlock(&vcpu->kvm->lock); + } return ret; } diff --git a/virt/kvm/arm/vgic/vgic-mmio-v3.c b/virt/kvm/arm/vgic/vgic-mmio-v3.c index 99da1a207c19..9b0f6810e7a8 100644 --- a/virt/kvm/arm/vgic/vgic-mmio-v3.c +++ b/virt/kvm/arm/vgic/vgic-mmio-v3.c @@ -586,7 +586,7 @@ int vgic_register_redist_iodev(struct kvm_vcpu *vcpu) if (!vgic_v3_check_base(kvm)) return -EINVAL; - rd_base = vgic->vgic_redist_base + kvm_vcpu_get_idx(vcpu) * SZ_64K * 2; + rd_base = vgic->vgic_redist_base + vgic->vgic_redist_free_offset; sgi_base = rd_base + SZ_64K; kvm_iodevice_init(&rd_dev->dev, &kvm_io_gic_ops); @@ -615,11 +615,14 @@ int vgic_register_redist_iodev(struct kvm_vcpu *vcpu) ret = kvm_io_bus_register_dev(kvm, KVM_MMIO_BUS, sgi_base, SZ_64K, &sgi_dev->dev); mutex_unlock(&kvm->slots_lock); - if (ret) + if (ret) { kvm_io_bus_unregister_dev(kvm, KVM_MMIO_BUS, &rd_dev->dev); + return ret; + } - return ret; + vgic->vgic_redist_free_offset += 2 * SZ_64K; + return 0; } static void vgic_unregister_redist_iodev(struct kvm_vcpu *vcpu) -- cgit v1.3-14-g43fede From 6a29beef9d1b16c762e469d77e28c3de3f5c3dbb Mon Sep 17 00:00:00 2001 From: Peter Chen Date: Wed, 17 May 2017 18:32:02 +0300 Subject: usb: host: xhci-ring: don't need to clear interrupt pending for MSI enabled hcd According to xHCI spec Figure 30: Interrupt Throttle Flow Diagram If PCI Message Signaled Interrupts (MSI or MSI-X) are enabled, then the assertion of the Interrupt Pending (IP) flag in Figure 30 generates a PCI Dword write. The IP flag is automatically cleared by the completion of the PCI write. the MSI enabled HCs don't need to clear interrupt pending bit, but hcd->irq = 0 doesn't equal to MSI enabled HCD. At some Dual-role controller software designs, it sets hcd->irq as 0 to avoid HCD requesting interrupt, and they want to decide when to call usb_hcd_irq by software. Signed-off-by: Peter Chen Signed-off-by: Mathias Nyman Signed-off-by: Greg Kroah-Hartman --- drivers/usb/host/xhci-ring.c | 5 +---- drivers/usb/host/xhci.c | 5 +++-- include/linux/usb/hcd.h | 1 + 3 files changed, 5 insertions(+), 6 deletions(-) (limited to 'include') diff --git a/drivers/usb/host/xhci-ring.c b/drivers/usb/host/xhci-ring.c index 507ba7734b94..0830b25f9499 100644 --- a/drivers/usb/host/xhci-ring.c +++ b/drivers/usb/host/xhci-ring.c @@ -2707,12 +2707,9 @@ irqreturn_t xhci_irq(struct usb_hcd *hcd) */ status |= STS_EINT; writel(status, &xhci->op_regs->status); - /* FIXME when MSI-X is supported and there are multiple vectors */ - /* Clear the MSI-X event interrupt status */ - if (hcd->irq) { + if (!hcd->msi_enabled) { u32 irq_pending; - /* Acknowledge the PCI interrupt */ irq_pending = readl(&xhci->ir_set->irq_pending); irq_pending |= IMAN_IP; writel(irq_pending, &xhci->ir_set->irq_pending); diff --git a/drivers/usb/host/xhci.c b/drivers/usb/host/xhci.c index 2d1310220832..71eb2991c698 100644 --- a/drivers/usb/host/xhci.c +++ b/drivers/usb/host/xhci.c @@ -359,9 +359,10 @@ static int xhci_try_enable_msi(struct usb_hcd *hcd) /* fall back to msi*/ ret = xhci_setup_msi(xhci); - if (!ret) - /* hcd->irq is 0, we have MSI */ + if (!ret) { + hcd->msi_enabled = 1; return 0; + } if (!pdev->irq) { xhci_err(xhci, "No msi-x/msi found and no IRQ in BIOS\n"); diff --git a/include/linux/usb/hcd.h b/include/linux/usb/hcd.h index a469999a106d..50398b69ca44 100644 --- a/include/linux/usb/hcd.h +++ b/include/linux/usb/hcd.h @@ -148,6 +148,7 @@ struct usb_hcd { unsigned rh_registered:1;/* is root hub registered? */ unsigned rh_pollable:1; /* may we poll the root hub? */ unsigned msix_enabled:1; /* driver has MSI-X enabled? */ + unsigned msi_enabled:1; /* driver has MSI enabled? */ unsigned remove_phy:1; /* auto-remove USB phy */ /* The next flag is a stopgap, to be removed when all the HCDs -- cgit v1.3-14-g43fede From 4123109050a869a8871e58a50f28f383d41e49ad Mon Sep 17 00:00:00 2001 From: James Smart Date: Fri, 5 May 2017 16:13:02 -0700 Subject: nvme-fc: correct port role bits FC Port roles is a bit mask, not individual values. Correct nvme definitions to unique bits. Signed-off-by: James Smart Reviewed-by: Johannes Thumshirn Signed-off-by: Christoph Hellwig Signed-off-by: Jens Axboe --- include/linux/nvme-fc-driver.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/include/linux/nvme-fc-driver.h b/include/linux/nvme-fc-driver.h index 0db37158a61d..12e344b5b77f 100644 --- a/include/linux/nvme-fc-driver.h +++ b/include/linux/nvme-fc-driver.h @@ -27,8 +27,8 @@ /* FC Port role bitmask - can merge with FC Port Roles in fc transport */ #define FC_PORT_ROLE_NVME_INITIATOR 0x10 -#define FC_PORT_ROLE_NVME_TARGET 0x11 -#define FC_PORT_ROLE_NVME_DISCOVERY 0x12 +#define FC_PORT_ROLE_NVME_TARGET 0x20 +#define FC_PORT_ROLE_NVME_DISCOVERY 0x40 /** -- cgit v1.3-14-g43fede From 4b8ba5fa525bc8bdaaed2a5c5433f0f2008d7bc5 Mon Sep 17 00:00:00 2001 From: James Smart Date: Tue, 25 Apr 2017 16:23:09 -0700 Subject: nvmet-fc: remove target cpu scheduling flag Remove NVMET_FCTGTFEAT_NEEDS_CMD_CPUSCHED. It's unnecessary. Signed-off-by: James Smart Reviewed-by: Johannes Thumshirn Signed-off-by: Christoph Hellwig Signed-off-by: Jens Axboe --- drivers/nvme/target/fc.c | 4 +--- drivers/nvme/target/fcloop.c | 1 - drivers/scsi/lpfc/lpfc_nvmet.c | 1 - include/linux/nvme-fc-driver.h | 12 ++---------- 4 files changed, 3 insertions(+), 15 deletions(-) (limited to 'include') diff --git a/drivers/nvme/target/fc.c b/drivers/nvme/target/fc.c index 62eba29c85fb..2006fae61980 100644 --- a/drivers/nvme/target/fc.c +++ b/drivers/nvme/target/fc.c @@ -517,9 +517,7 @@ nvmet_fc_queue_to_cpu(struct nvmet_fc_tgtport *tgtport, int qid) { int cpu, idx, cnt; - if (!(tgtport->ops->target_features & - NVMET_FCTGTFEAT_NEEDS_CMD_CPUSCHED) || - tgtport->ops->max_hw_queues == 1) + if (tgtport->ops->max_hw_queues == 1) return WORK_CPU_UNBOUND; /* Simple cpu selection based on qid modulo active cpu count */ diff --git a/drivers/nvme/target/fcloop.c b/drivers/nvme/target/fcloop.c index 15551ef79c8c..294a6611fb24 100644 --- a/drivers/nvme/target/fcloop.c +++ b/drivers/nvme/target/fcloop.c @@ -698,7 +698,6 @@ static struct nvmet_fc_target_template tgttemplate = { .dma_boundary = FCLOOP_DMABOUND_4G, /* optional features */ .target_features = NVMET_FCTGTFEAT_CMD_IN_ISR | - NVMET_FCTGTFEAT_NEEDS_CMD_CPUSCHED | NVMET_FCTGTFEAT_OPDONE_IN_ISR, /* sizes of additional private data for data structures */ .target_priv_sz = sizeof(struct fcloop_tport), diff --git a/drivers/scsi/lpfc/lpfc_nvmet.c b/drivers/scsi/lpfc/lpfc_nvmet.c index 94434e621c33..0488580eea12 100644 --- a/drivers/scsi/lpfc/lpfc_nvmet.c +++ b/drivers/scsi/lpfc/lpfc_nvmet.c @@ -764,7 +764,6 @@ lpfc_nvmet_create_targetport(struct lpfc_hba *phba) lpfc_tgttemplate.max_sgl_segments = phba->cfg_nvme_seg_cnt + 1; lpfc_tgttemplate.max_hw_queues = phba->cfg_nvme_io_channel; lpfc_tgttemplate.target_features = NVMET_FCTGTFEAT_READDATA_RSP | - NVMET_FCTGTFEAT_NEEDS_CMD_CPUSCHED | NVMET_FCTGTFEAT_CMD_IN_ISR | NVMET_FCTGTFEAT_OPDONE_IN_ISR; diff --git a/include/linux/nvme-fc-driver.h b/include/linux/nvme-fc-driver.h index 12e344b5b77f..6c8c5d8041b7 100644 --- a/include/linux/nvme-fc-driver.h +++ b/include/linux/nvme-fc-driver.h @@ -642,15 +642,7 @@ enum { * sequence in one LLDD operation. Errors during Data * sequence transmit must not allow RSP sequence to be sent. */ - NVMET_FCTGTFEAT_NEEDS_CMD_CPUSCHED = (1 << 1), - /* Bit 1: When 0, the LLDD will deliver FCP CMD - * on the CPU it should be affinitized to. Thus work will - * be scheduled on the cpu received on. When 1, the LLDD - * may not deliver the CMD on the CPU it should be worked - * on. The transport should pick a cpu to schedule the work - * on. - */ - NVMET_FCTGTFEAT_CMD_IN_ISR = (1 << 2), + NVMET_FCTGTFEAT_CMD_IN_ISR = (1 << 1), /* Bit 2: When 0, the LLDD is calling the cmd rcv handler * in a non-isr context, allowing the transport to finish * op completion in the calling context. When 1, the LLDD @@ -658,7 +650,7 @@ enum { * requiring the transport to transition to a workqueue * for op completion. */ - NVMET_FCTGTFEAT_OPDONE_IN_ISR = (1 << 3), + NVMET_FCTGTFEAT_OPDONE_IN_ISR = (1 << 2), /* Bit 3: When 0, the LLDD is calling the op done handler * in a non-isr context, allowing the transport to finish * op completion in the calling context. When 1, the LLDD -- cgit v1.3-14-g43fede