aboutsummaryrefslogtreecommitdiffstats
path: root/drivers/gpu/drm/mgag200 (follow)
AgeCommit message (Collapse)AuthorFilesLines
2021-08-18drm/mgag200: Fix uninitialized variable deltaColin Ian King1-0/+1
The variable delta is not initialized and this will cause unexpected behaviour with the comparison of tmpdelta < delta. Fix this by setting it to 0xffffffff. This matches the behaviour as in the similar function mgag200_pixpll_compute_g200se_04. v2: * move fix up by one line to align style with other functions * add additional tags from similar patch Signed-off-by: Colin Ian King <colin.king@canonical.com> Fixes: 2545ac960364 ("drm/mgag200: Abstract pixel PLL via struct mgag200_pll") Addresses-Coverity: ("Uninitialized scalar variable") Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch> Reviewed-by: Thomas Zimmermann <tzimmermann@suse.de> Reported-by: kernel test robot <lkp@intel.com> Reported-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Dave Airlie <airlied@redhat.com> Cc: dri-devel@lists.freedesktop.org Link: https://patchwork.freedesktop.org/patch/msgid/20210817163204.494166-1-colin.king@canonical.com
2021-08-08drm/mgag200: Use offset-adjusted shadow-plane mappingsThomas Zimmermann1-2/+2
For framebuffers with non-zero offset fields, shadow-plane helpers provide a pointer to the first byte of the contained data. Use it in mgag200. Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Acked-by: Sam Ravnborg <sam@ravnborg.org> Link: https://patchwork.freedesktop.org/patch/msgid/20210803125928.27780-6-tzimmermann@suse.de
2021-08-08drm/mgag200: Compute PLL values during atomic checkThomas Zimmermann2-18/+22
PLL setup can fail if the display mode's clock is not supported by any PLL configuration. Compute the PLL values during atomic check, so that atomic commits can fail at the appropriate time. If successful, use the values in the atomic-update phase. Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Acked-by: Sam Ravnborg <sam@ravnborg.org> Link: https://patchwork.freedesktop.org/patch/msgid/20210714142240.21979-14-tzimmermann@suse.de
2021-08-08drm/mgag200: Introduce custom CRTC stateThomas Zimmermann2-0/+55
Inherit from struct drm_crtc_state by embeding it and providing the rsp callbacks for simple-kms helpers. No functional changes. The new state struct mgag200_crtc_state will hold PLL values for modeset operations. v2: * move the simple-kms changes into a separate patch (Sam) Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Acked-by: Sam Ravnborg <sam@ravnborg.org> Link: https://patchwork.freedesktop.org/patch/msgid/20210714142240.21979-13-tzimmermann@suse.de
2021-08-08drm/mgag200: Abstract pixel PLL via struct mgag200_pllThomas Zimmermann4-903/+1019
Move all PLL compute and update functions into mgag200_pll.c. No functional changes to the rsp algorithms. Introduce struct mgag200_pll and mgag200_pll_funcs. The data strutures abstract the details of each revision's PLL. Perform calls to compute and update functionality via function pointers. Init the PLL once as part of the driver initialization. Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Acked-by: Sam Ravnborg <sam@ravnborg.org> Link: https://patchwork.freedesktop.org/patch/msgid/20210714142240.21979-11-tzimmermann@suse.de
2021-08-08drm/mgag200: Declare PLL clock constants static constThomas Zimmermann1-44/+31
Move the PLL constants to the RO data section by declaring them as static const. No functional changes. Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Acked-by: Sam Ravnborg <sam@ravnborg.org> Link: https://patchwork.freedesktop.org/patch/msgid/20210714142240.21979-10-tzimmermann@suse.de
2021-08-08drm/mgag200: Split PLL compute function for G200SE by revThomas Zimmermann1-70/+96
The compute function for G200SE pixel PLLs handles two revisions with different algorithms. Split it accordingly to make it readable. No functional changes. Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Acked-by: Sam Ravnborg <sam@ravnborg.org> Link: https://patchwork.freedesktop.org/patch/msgid/20210714142240.21979-9-tzimmermann@suse.de
2021-08-08drm/mgag200: Split PLL compute functions by device typeThomas Zimmermann1-110/+146
Several PLL functions compute values for different device types. Split them up to make the code more readable. No functional changes. Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Acked-by: Sam Ravnborg <sam@ravnborg.org> Link: https://patchwork.freedesktop.org/patch/msgid/20210714142240.21979-8-tzimmermann@suse.de
2021-08-08drm/mgag200: Store values (not bits) in struct mgag200_pll_valuesThomas Zimmermann1-62/+91
The fields in struct mgag200_pll_values currently hold the bits of each register. Store the PLL values instead and let the PLL-update code figure out the bits for each register. Until now, the compute function either stored plain values or register bits in struct mgag200_pll_values. The rsp update function used the values as-is. This made it very hard to correctly interpret the stored values (e.g., for logging or debugging). With the cleanup, the stored values now have a clear meaning. v2: * add a bit more context in the commit message (Sam) Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Acked-by: Sam Ravnborg <sam@ravnborg.org> Link: https://patchwork.freedesktop.org/patch/msgid/20210714142240.21979-7-tzimmermann@suse.de
2021-08-08drm/mgag200: Introduce separate variable for PLL S parameterThomas Zimmermann1-20/+16
The S parameter is controls the loop filter bandwidth when programming the PLL. It's currently stored as part of P (i.e., the clock divider.) Add a separate variable for S prepares the PLL code for further refactoring. The value of s is currently 0, so it has not yet an effect on the programming. v2: * add a note on the current value of s to commit message Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Acked-by: Sam Ravnborg <sam@ravnborg.org> Link: https://patchwork.freedesktop.org/patch/msgid/20210714142240.21979-6-tzimmermann@suse.de
2021-08-08drm/mgag200: Split PLL setup into compute and update functionsThomas Zimmermann2-52/+198
The _set_plls() functions compute a pixel clock's PLL values and program the hardware accordingly. This happens during atomic commits. For atomic modesetting, it's better to separate computation and programming from each other. This will allow to compute the PLL value during atomic checks and catch unsupported modes early. Split the PLL setup into a compute and an update functions, and call them one after the other. Computed PLL values are store in struct mgag200_pll_values. There are four parameters for the PLL, m, n, p and s. Every compute function stores a value for each of these parameters, and the rsp update function makes the register bits from them. The values stored by the compute function are either plain values or register bits. An additional change is required to always store plain values. No functional changes. Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Acked-by: Sam Ravnborg <sam@ravnborg.org> Link: https://patchwork.freedesktop.org/patch/msgid/20210714142240.21979-5-tzimmermann@suse.de
2021-08-08drm/mgag200: Remove P_ARRAY_SIZEThomas Zimmermann1-4/+3
Replace P_ARRAY_SIZE by array pre-initializing and ARRAY_SIZE(). No functional changes. Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Acked-by: Sam Ravnborg <sam@ravnborg.org> Link: https://patchwork.freedesktop.org/patch/msgid/20210714142240.21979-4-tzimmermann@suse.de
2021-08-08drm/mgag200: Return errno codes from PLL compute functionsThomas Zimmermann1-2/+2
Return -EINVAL if there's no PLL configuration for the given pixel clock. The returned errors are currently ignored by the caller, but the errno codes will become useful when the compute functions run during atomic checks. v2: * give a rational for this change (Sam) Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Acked-by: Sam Ravnborg <sam@ravnborg.org> Link: https://patchwork.freedesktop.org/patch/msgid/20210714142240.21979-3-tzimmermann@suse.de
2021-08-08drm/mgag200: Select clock in PLL update functionsThomas Zimmermann3-12/+33
Put the clock-selection code into each of the PLL-update functions to make them select the correct pixel clock. Instead of copying the code, introduce a new helper WREG_MISC_MASKED, which does masked writes into <MISC>. Use it from each individual PLL update function. The pixel clock for video output was not actually set before programming the clock's values. It worked because the device had the correct clock pre-set. v2: * don't duplicate <MISC> update code (Sam) Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Fixes: db05f8d3dc87 ("drm/mgag200: Split MISC register update into PLL selection, SYNC and I/O") Acked-by: Sam Ravnborg <sam@ravnborg.org> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Emil Velikov <emil.velikov@collabora.com> Cc: Dave Airlie <airlied@redhat.com> Cc: dri-devel@lists.freedesktop.org Cc: <stable@vger.kernel.org> # v5.9+ Link: https://patchwork.freedesktop.org/patch/msgid/20210714142240.21979-2-tzimmermann@suse.de
2021-07-05drm/mgag200: Constify LUT for programming bppThomas Zimmermann2-12/+6
Declare constant LUT for bpp programming as static const. Removes mutable data from device structure. Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Acked-by: Sam Ravnborg <sam@ravnborg.org> Link: https://patchwork.freedesktop.org/patch/msgid/20210702075642.27834-5-tzimmermann@suse.de
2021-07-05drm/mgag200: Extract device type and flags in mgag200_pci_probe()Thomas Zimmermann2-16/+17
The type and flags values are stored in the PCI ID list. Extract them in the probe function. Makes the device initialization more readable. Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Acked-by: Sam Ravnborg <sam@ravnborg.org> Link: https://patchwork.freedesktop.org/patch/msgid/20210702075642.27834-4-tzimmermann@suse.de
2021-07-05drm/mgag200: Inline mgag200_device_init()Thomas Zimmermann1-28/+13
Inline mgag200_device_init() into mgag200_device_create(), which is the only caller. Also remove a duplicate error message for mgag200_modeset_init(). The function will print an error if/where it failed. v2: * include a note about the removed error message in changelog (Sam) Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Acked-by: Sam Ravnborg <sam@ravnborg.org> Link: https://patchwork.freedesktop.org/patch/msgid/20210702075642.27834-3-tzimmermann@suse.de
2021-07-05drm/mgag200: Don't pass flags to drm_dev_register()Thomas Zimmermann1-1/+1
The flags argument is only relevant for UMS drivers. Pass 0. Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Acked-by: Sam Ravnborg <sam@ravnborg.org> Link: https://patchwork.freedesktop.org/patch/msgid/20210702075642.27834-2-tzimmermann@suse.de
2021-07-01drm/aperture: Pass DRM driver structure instead of driver nameThomas Zimmermann1-1/+1
Print the name of the DRM driver when taking over fbdev devices. Makes the output to dmesg more consistent. Note that the driver name is only used for printing a string to the kernel log. No UAPI is affected by this change. Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Acked-by: Nirmoy Das <nirmoy.das@amd.com> Acked-by: Chen-Yu Tsai <wens@csie.org> # sun4i Acked-by: Neil Armstrong <narmstrong@baylibre.com> # meson Link: https://patchwork.freedesktop.org/patch/msgid/20210629135833.22679-1-tzimmermann@suse.de
2021-05-01drm/format-helper: Pass destination pitch to drm_fb_memcpy_dstclip()Thomas Zimmermann1-1/+1
The memcpy's destination buffer might have a different pitch than the source. Support different pitches as function argument. Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch> Acked-by: Maxime Ripard <maxime@cerno.tech> Tested-by: nerdopolis <bluescreen_avenger@verizon.net> Link: https://patchwork.freedesktop.org/patch/msgid/20210430105840.30515-2-tzimmermann@suse.de
2021-04-14drm/aperture: Convert drivers to aperture interfacesThomas Zimmermann1-1/+4
Mass-convert all drivers from FB helpers to aperture interfaces. No functional changes besides checking for returned errno codes. Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Acked-by: Jani Nikula <jani.nikula@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20210412131043.5787-3-tzimmermann@suse.de
2021-02-08drm/mgag200: Move vmap out of commit tailThomas Zimmermann1-15/+8
Vmap operations may acquire the dmabuf reservation lock, which is not allowed within atomic commit-tail functions. Therefore move vmap and vunmap from the damage handler into prepare_fb and cleanup_fb callbacks. The mapping is provided as GEM shadow-buffered plane. The functions in the commit tail use the pre-established mapping for damage handling. Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Tested-by: Gerd Hoffmann <kraxel@redhat.com> Acked-by: Gerd Hoffmann <kraxel@redhat.com> Link: https://patchwork.freedesktop.org/patch/msgid/20210208115538.6430-5-tzimmermann@suse.de
2021-02-05drm/mgag200: make a const array static, makes object smallerColin Ian King1-1/+1
Don't populate the const array m_div_val on the stack but instead make it static. Makes the object code smaller by 29 bytes: Before: text data bss dec hex filename 34736 4552 0 39288 9978 drivers/gpu/drm/mgag200/mgag200_mode.o After: text data bss dec hex filename 34625 4616 0 39241 9949 drivers/gpu/drm/mgag200/mgag200_mode.o (gcc version 10.2.0) Signed-off-by: Colin Ian King <colin.king@canonical.com> Reviewed-by: Thomas Zimmermann <tzimmermann@suse.de> Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Link: https://patchwork.freedesktop.org/patch/msgid/20210204191156.110778-1-colin.king@canonical.com
2021-01-06drm/mgag200: Remove references to struct drm_device.pdevThomas Zimmermann3-14/+18
Using struct drm_device.pdev is deprecated. Convert mgag200 to struct drm_device.dev. No functional changes. Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Acked-by: Sam Ravnborg <sam@ravnborg.org> Link: https://patchwork.freedesktop.org/patch/msgid/20201201103542.2182-13-tzimmermann@suse.de
2020-11-24drm/shmem-helper: Removed drm_gem_shmem_create_object_cached()Thomas Zimmermann1-1/+0
Cached page mappings are now the default for SHMEM GEM objects. Remove the obsolete create function for cached mappings. Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Acked-by: Maxime Ripard <mripard@kernel.org> Link: https://patchwork.freedesktop.org/patch/msgid/20201117133156.26822-3-tzimmermann@suse.de
2020-11-09drm/gem: Use struct dma_buf_map in GEM vmap ops and convert GEM backendsThomas Zimmermann1-3/+7
This patch replaces the vmap/vunmap's use of raw pointers in GEM object functions with instances of struct dma_buf_map. GEM backends are converted as well. For most of them, this simply changes the returned type. TTM-based drivers now return information about the location of the memory, either system or I/O memory. GEM VRAM helpers and qxl now use ttm_bo_vmap() et al. Amdgpu, nouveau and radeon use drm_gem_ttm_vmap() et al instead of implementing their own vmap callbacks. v7: * init QXL cursor to mapped BO buffer (kernel test robot) v5: * update vkms after switch to shmem v4: * use ttm_bo_vmap(), drm_gem_ttm_vmap(), et al. (Daniel, Christian) * fix a trailing { in drm_gem_vmap() * remove several empty functions instead of converting them (Daniel) * comment uses of raw pointers with a TODO (Daniel) * TODO list: convert more helpers to use struct dma_buf_map Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Acked-by: Christian König <christian.koenig@amd.com> Tested-by: Sam Ravnborg <sam@ravnborg.org> Link: https://patchwork.freedesktop.org/patch/msgid/20201103093015.1063-7-tzimmermann@suse.de
2020-11-08drm: remove unneeded breakTom Rix1-5/+0
A break is not needed if it is preceded by a return or break Signed-off-by: Tom Rix <trix@redhat.com> Acked-by: Sam Ravnborg <sam@ravnborg.org> Signed-off-by: Sam Ravnborg <sam@ravnborg.org> Link: https://patchwork.freedesktop.org/patch/msgid/20201019163115.25814-1-trix@redhat.com
2020-11-06drm/<drivers>: Constify struct drm_driverDaniel Vetter1-1/+1
Only the following drivers aren't converted: - amdgpu, because of the driver_feature mangling due to virt support. Subsequent patch will address this. - nouveau, because DRIVER_ATOMIC uapi is still not the default on the platforms where it's supported (i.e. again driver_feature mangling) - vc4, again because of driver_feature mangling - qxl, because the ioctl table is somewhere else and moving that is maybe a bit too much, hence the num_ioctls assignment prevents a const driver structure. - arcpgu, because that is stuck behind a pending tiny-fication series from me. - legacy drivers, because legacy requires non-const drm_driver. Note that for armada I also went ahead and made the ioctl array const. Only cc'ing the driver people who've not been converted (everyone else is way too much). v2: Fix one misplaced const static, should be static const (0day) v3: - Improve commit message (Sam) Acked-by: Sam Ravnborg <sam@ravnborg.org> Cc: kernel test robot <lkp@intel.com> Acked-by: Maxime Ripard <mripard@kernel.org> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Dave Airlie <airlied@redhat.com> Cc: Gerd Hoffmann <kraxel@redhat.com> Cc: virtualization@lists.linux-foundation.org Cc: Harry Wentland <harry.wentland@amd.com> Cc: Leo Li <sunpeng.li@amd.com> Cc: Alex Deucher <alexander.deucher@amd.com> Cc: Christian König <christian.koenig@amd.com> Cc: Eric Anholt <eric@anholt.net> Cc: Maxime Ripard <mripard@kernel.org> Cc: Ben Skeggs <bskeggs@redhat.com> Cc: nouveau@lists.freedesktop.org Signed-off-by: Daniel Vetter <daniel.vetter@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20201104100425.1922351-5-daniel.vetter@ffwll.ch
2020-08-27drm/mgag200: fix spelling mistake "expeced" -> "expected"Colin Ian King1-1/+1
There is a spelling mistake in a drm_warn message. Fix it. Signed-off-by: Colin Ian King <colin.king@canonical.com> Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Link: https://patchwork.freedesktop.org/patch/msgid/20200826084727.42703-1-colin.king@canonical.com
2020-08-07drm/mgag200: fix build on alpha archSam Ravnborg1-0/+1
When building imgag200 for the alpha architecture it fails like this: mgag200_drv.c:233:9: error: implicit declaration of function ‘vmalloc’ 233 | bios = vmalloc(size); | ^~~~~~~ | kmalloc When building for other architectures vmalloc.h is pulled in via some other header file - for example asm-generic/io.h. Use an explicit include of vmalloc.h to fix the build. Signed-off-by: Sam Ravnborg <sam@ravnborg.org> Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch> Fixes: e20dfd27f7aa ("drm/mgag200: Add support for G200 desktop cards") Cc: Thomas Zimmermann <tzimmermann@suse.de> Cc: Egbert Eich <eich@suse.com> Cc: Takashi Iwai <tiwai@suse.de> Cc: Lyude Paul <lyude@redhat.com> Link: https://patchwork.freedesktop.org/patch/msgid/20200807180547.GA923146@ravnborg.org
2020-08-05drm/mgag200: Set PCI option register in G200SE modelsThomas Zimmermann1-0/+1
The initial value of the PCI option register got lost while refactoring the driver init code. Restore the setting. Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Reviewed-by: Dave Airlie <airlied@redhat.com> Reported-by: kernel test robot <lkp@intel.com> Fixes: 2021708e0d6e ("drm/mgag200: Initialize PCI registers early during device setup") Cc: Lyude Paul <lyude@redhat.com> Cc: Dave Airlie <airlied@redhat.com> Cc: Sam Ravnborg <sam@ravnborg.org> Cc: Daniel Vetter <daniel.vetter@ffwll.ch> Cc: Gerd Hoffmann <kraxel@redhat.com> Cc: Emil Velikov <emil.velikov@collabora.com> Link: https://patchwork.freedesktop.org/patch/msgid/20200804065158.21049-1-tzimmermann@suse.de
2020-08-03drm/mgag200: Add support for G200 desktop cardsThomas Zimmermann4-8/+231
This patch adds support for G200 desktop cards. We can reuse the whole memory and modesetting code. A few PCI and DAC register values have to be updated accordingly. The most significant change is in the PLL setup. The driver parses the device's BIOS to retrieve clock limits and reference clocks. With no BIOS found, safe defaults are being used. v2: * copy BIOS ROM to system memory and access with regular load/store; resolves potential HW limitations * fix some stray whitespaces Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Co-developed-by: Egbert Eich <eich@suse.com> Signed-off-by: Egbert Eich <eich@suse.com> Co-developed-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Takashi Iwai <tiwai@suse.de> Reviewed-by: Lyude Paul <lyude@redhat.com> Link: https://patchwork.freedesktop.org/patch/msgid/20200730102844.10995-9-tzimmermann@suse.de
2020-08-03drm/mgag200: Move G200SE's unique id into model-specific dataThomas Zimmermann3-15/+30
The unique revision id is only useful for G200SE devices. Store the value in model-specific data within struct mga_device. While at it, the patch also adds an init helper for the value. Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Reviewed-by: Lyude Paul <lyude@redhat.com> Link: https://patchwork.freedesktop.org/patch/msgid/20200730102844.10995-8-tzimmermann@suse.de
2020-08-03drm/mgag200: Clear <page> field during MM initThomas Zimmermann2-5/+3
The modesetting code initialized the memory-related register CRTCEXT4. Move this code to MM initialization. Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Reviewed-by: Lyude Paul <lyude@redhat.com> Link: https://patchwork.freedesktop.org/patch/msgid/20200730102844.10995-7-tzimmermann@suse.de
2020-08-03drm/mgag200: Set MISC memory flags in mm init codeThomas Zimmermann2-3/+7
The modesetting code initialized several memory-related flags in the MISC register. Move this code to MM initialization. Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Reviewed-by: Lyude Paul <lyude@redhat.com> Link: https://patchwork.freedesktop.org/patch/msgid/20200730102844.10995-6-tzimmermann@suse.de
2020-08-03drm/mgag200: Enable MGA mode during device register initializationThomas Zimmermann3-5/+8
MGA cards can run in traditional VGA mode or an enhanced MGA mode; with the latter being required for KMS. So far, MGA mode was enabled during modesetting. As it's fundamental for device operation, the patch moves it next to the device register setup. Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Reviewed-by: Lyude Paul <lyude@redhat.com> Link: https://patchwork.freedesktop.org/patch/msgid/20200730102844.10995-5-tzimmermann@suse.de
2020-08-03drm/mgag200: Initialize PCI registers early during device setupThomas Zimmermann3-43/+31
So far, PCI option registers were initialized as part of modesetting, which is late in the process. As these registers control fundamental operation, they should be set early. The patch moves the PCI option handling into device register setup, before even the device MMIO memory is being mapped. No functional changes made. Moving the PCI code next to the device-register setup also allows to remove the has_sdram field from struct mga_device. The state is now local to the init helper. Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Reviewed-by: Lyude Paul <lyude@redhat.com> Link: https://patchwork.freedesktop.org/patch/msgid/20200730102844.10995-4-tzimmermann@suse.de
2020-08-03drm/mgag200: Move register initialization into helper functionThomas Zimmermann2-7/+32
The mgag200 driver maps registers into the address space. Move the code into a separate helper function. No functional changes. One small difference is in the handling of SDRAM/SGRAM. MGA devices can come with either SDRAM or SGRAM. So far, the driver checked for SDRAM, which is the common case. The patch moves this code into a separate helper and checks for SGRAM, which is the special case. The test itself is the same as before. Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Reviewed-by: Lyude Paul <lyude@redhat.com> Link: https://patchwork.freedesktop.org/patch/msgid/20200730102844.10995-3-tzimmermann@suse.de
2020-08-03drm/mgag200: Enable caching for SHMEM pagesThomas Zimmermann1-0/+1
SHMEM pages use write-combine caching by default, but can also use the platform's default page caching. Doing so may improve the performance of I/O on the framebuffer. Mgag200's hardware does not access framebuffer pages directly (i.e., via DMA), so enabling caching does not have an effect on consistency of the framebuffer memory or the displayed data. Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Reviewed-by: Lyude Paul <lyude@redhat.com> Link: https://patchwork.freedesktop.org/patch/msgid/20200730102844.10995-2-tzimmermann@suse.de
2020-07-14drm/mgag200: Inline mga_crtc_{prepare, commit}() into enable functionThomas Zimmermann1-32/+7
There's only trivial code left in mga_crtc_{prepare,commit}(). Merge the functions into the simple pipe's enable function. Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Reviewed-by: Lyude Paul <lyude@redhat.com> Link: https://patchwork.freedesktop.org/patch/msgid/20200707082411.6583-8-tzimmermann@suse.de
2020-07-14drm/mgag200: Rename G200WB prepare/commit functionThomas Zimmermann1-6/+4
The prepare and commit helpers for G200WB devices control the BMC. Rename them accordingly. While at it, also change the passed value's type to struct mga_device and remove some type upcasting. Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Reviewed-by: Lyude Paul <lyude@redhat.com> Link: https://patchwork.freedesktop.org/patch/msgid/20200707082411.6583-7-tzimmermann@suse.de
2020-07-14drm/mgag200: Set/clear <syncrst> field in display enable/disable helpersThomas Zimmermann2-6/+14
Modifying the <syncrst> field in mgag200_{enable,disable}_display() makes the code more readable. Also clear the <asyncrst> field to enable the display. The other bits in SEQ0 are unused, so no functional changes are made. Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Reviewed-by: Lyude Paul <lyude@redhat.com> Link: https://patchwork.freedesktop.org/patch/msgid/20200707082411.6583-6-tzimmermann@suse.de
2020-07-14drm/mgag200: Split DPMS function into helpersThomas Zimmermann2-31/+45
Of the DPMS code, only ON and OFF states are used. Simplify mode setting by moving both into separate functions and removing the rest. The original code busy waited in the middle of updating the screen state in SEQ1. To simplify the procedure, the new code busy waits first and then updates SEQ1 in one chunk. The DPMS code also set the LUT before enabling the screen. The patch moves this code into the simple-display pipe's enable function. v2: * comment on SEQ1 updates in commit message Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Reviewed-by: Lyude Paul <lyude@redhat.com> Link: https://patchwork.freedesktop.org/patch/msgid/20200707082411.6583-5-tzimmermann@suse.de
2020-07-14drm/mgag200: Don't set or clear <scroff> field during modesetThomas Zimmermann1-39/+3
The simple pipe's disable function disables the screen by calling mgag200_disable_screen(). The simple pipe's enable function enables the screen by calling mgag200_enable_display(). During modeset operations the screen is off and remains off. It's only enabled after the modeset has been completed. Therefore remove all code that sets or clears the <scroff> field while in modeset. The related code also modifies the <syncrst> field in SEQ0. For now, keep this code in place. Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Reviewed-by: Lyude Paul <lyude@redhat.com> Link: https://patchwork.freedesktop.org/patch/msgid/20200707082411.6583-4-tzimmermann@suse.de
2020-07-14drm/mgag200: Move PLL setup out of mode-setting functionThomas Zimmermann1-3/+2
Makes the code slightly more flexible. No functional changes. Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Reviewed-by: Lyude Paul <lyude@redhat.com> Link: https://patchwork.freedesktop.org/patch/msgid/20200707082411.6583-3-tzimmermann@suse.de
2020-07-14drm/mgag200: Don't write-protect CRTC 0-7 while in mga_crtc_prepare()Thomas Zimmermann3-9/+18
The prepare function write-protects several registers that it doesn't even touch. Removed the related code. The code for unprotecting registers also clears VINT interrupts. Both is now done once during initialization. Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Reviewed-by: Lyude Paul <lyude@redhat.com> Link: https://patchwork.freedesktop.org/patch/msgid/20200707082411.6583-2-tzimmermann@suse.de
2020-06-11drm/mgag200: Use managed device initializationThomas Zimmermann1-22/+8
The mgag200 driver now uses managed functions for DRM devices. The individual helpers for modesetting and memory managed are already covered, so only device allocation and initialization is left for conversion. Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Acked-by: Sam Ravnborg <sam@ravnborg.org> Link: https://patchwork.freedesktop.org/patch/msgid/20200605135803.19811-15-tzimmermann@suse.de
2020-06-11drm/mgag200: Embed instance of struct drm_device in struct mga_deviceThomas Zimmermann4-38/+27
Following current best practice, the instance of struct drm_device is now embedded in struct mga_device. The respective field has been renamed from 'dev' to 'base' to reflect the relationship. Conversion from DRM device is done via upcast. Using dev_private is no longer possible. The patch also open-codes drm_dev_alloc() and DRM device initialization is now performed by a call to drm_device_init(). Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Acked-by: Sam Ravnborg <sam@ravnborg.org> Link: https://patchwork.freedesktop.org/patch/msgid/20200605135803.19811-14-tzimmermann@suse.de
2020-06-11drm/mgag200: Allocate device structures in mgag200_driver_load()Thomas Zimmermann1-17/+21
Instances of struct drm_device and struct mga_device are now allocated next to each other in mgag200_driver_load(). Yet another preparation before embedding the DRM device instance in struct mga_device. Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Acked-by: Sam Ravnborg <sam@ravnborg.org> Link: https://patchwork.freedesktop.org/patch/msgid/20200605135803.19811-13-tzimmermann@suse.de
2020-06-11drm/mgag200: Separate device initialization into allocationThomas Zimmermann1-10/+22
Embedding the DRM device instance in struct mga_device will require changes to device allocation. Moving the device initialization into its own functions gets it out of the way. Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Acked-by: Sam Ravnborg <sam@ravnborg.org> Link: https://patchwork.freedesktop.org/patch/msgid/20200605135803.19811-12-tzimmermann@suse.de
Copyright © 1996 – 2023 Jason A. Donenfeld. All Rights Reverse Engineered.