Most of our helpers have relied on being selected so far through
Kconfig, but that creates issues when we have multiple layers of helpers
with some depending on others.
Indeed, select doesn't select a dependency's dependencies, and thus
isn't super intuitive. Depends on however doesn't have that limitation,
so we can just switch all the drivers that were selecting
DRM_DISPLAY_HDMI_HELPER to depend on it.
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Link: https://lore.kernel.org/r/20240327-kms-kconfig-helpers-v3-12-eafee11b84b3@kernel.org
Signed-off-by: Maxime Ripard <mripard@kernel.org>
Most of our helpers have relied on being selected so far through
Kconfig, but that creates issues when we have multiple layers of helpers
with some depending on others.
Indeed, select doesn't select a dependency's dependencies, and thus
isn't super intuitive. Depends on however doesn't have that limitation,
so we can just switch all the drivers that were selecting
DRM_DISPLAY_HDCP_HELPER to depend on it.
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Link: https://lore.kernel.org/r/20240327-kms-kconfig-helpers-v3-11-eafee11b84b3@kernel.org
Signed-off-by: Maxime Ripard <mripard@kernel.org>
Most of our helpers have relied on being selected so far through
Kconfig, but that creates issues when we have multiple layers of helpers
with some depending on others.
Indeed, select doesn't select a dependency's dependencies, and thus
isn't super intuitive. Depends on however doesn't have that limitation,
so we can just switch all the drivers that were selecting
DRM_DISPLAY_DP_HELPER to depend on it.
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Link: https://lore.kernel.org/r/20240327-kms-kconfig-helpers-v3-10-eafee11b84b3@kernel.org
Signed-off-by: Maxime Ripard <mripard@kernel.org>
Most of our helpers have relied on being selected so far through
Kconfig, but that creates issues when we have multiple layers of helpers
with some depending on others.
Indeed, select doesn't select a dependency's dependencies, and thus
isn't super intuitive. Depends on however doesn't have that limitation,
so we can just switch all the drivers that were selecting
DRM_DISPLAY_DP_AUX_BUS to depend on it.
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Link: https://lore.kernel.org/r/20240327-kms-kconfig-helpers-v3-9-eafee11b84b3@kernel.org
Signed-off-by: Maxime Ripard <mripard@kernel.org>
Most of our helpers have relied on being selected so far through
Kconfig, but that creates issues when we have multiple layers of helpers
with some depending on others.
Indeed, select doesn't select a dependency's dependencies, and thus
isn't super intuitive. Depends on however doesn't have that limitation,
so we can just switch all the drivers that were selecting
DRM_DISPLAY_HELPER to depend on it.
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Link: https://lore.kernel.org/r/20240327-kms-kconfig-helpers-v3-8-eafee11b84b3@kernel.org
Signed-off-by: Maxime Ripard <mripard@kernel.org>
DRM_DW_HDMI has a number of dependencies that might not be enabled.
However, drivers were used to selecting it while not enforcing the
DRM_DW_HDMI dependencies.
This could result in Kconfig warnings (and further build breakages) such
as:
Kconfig warnings: (for reference only)
WARNING: unmet direct dependencies detected for DRM_DW_HDMI
Depends on [n]: HAS_IOMEM [=y] && DRM [=m] && DRM_BRIDGE [=y] && DRM_DISPLAY_HELPER [=n]
Selected by [m]:
- DRM_SUN8I_DW_HDMI [=m] && HAS_IOMEM [=y] && DRM_SUN4I [=m]
Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202403262127.kZkttfNz-lkp@intel.com/
Acked-by: Jani Nikula <jani.nikula@intel.com>
Link: https://lore.kernel.org/r/20240327-kms-kconfig-helpers-v3-7-eafee11b84b3@kernel.org
Signed-off-by: Maxime Ripard <mripard@kernel.org>
All the helpers Kconfig symbols so far have relied on drivers selecting
them, and that's what most drivers did.
However, this creates an issue nowadays when helpers depend on each
other, and select doesn't transitively select a dependency dependencies.
Depends on doesn't have that limitation though, so let's convert those
symbols to be dependable and use depends on between them too.
Suggested-by: Jani Nikula <jani.nikula@linux.intel.com>
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Link: https://lore.kernel.org/r/20240327-kms-kconfig-helpers-v3-6-eafee11b84b3@kernel.org
Signed-off-by: Maxime Ripard <mripard@kernel.org>
The display kconfig helpers are not organized in any particular order,
so let's move them around to create an alphabetical order.
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Reviewed-by: Lucas De Marchi <lucas.demarchi@intel.com>
Link: https://lore.kernel.org/r/20240327-kms-kconfig-helpers-v3-5-eafee11b84b3@kernel.org
Signed-off-by: Maxime Ripard <mripard@kernel.org>
While most display helpers Kconfig symbols have the DRM_DISPLAY prefix,
the DisplayPort CEC tunnelling implementation uses CONFIG_DRM_DISPLAY_DP_AUX_CEC.
Since the number of users is limited, we can easily rename it to make it
consistent.
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Reviewed-by: Lucas De Marchi <lucas.demarchi@intel.com>
Link: https://lore.kernel.org/r/20240327-kms-kconfig-helpers-v3-4-eafee11b84b3@kernel.org
Signed-off-by: Maxime Ripard <mripard@kernel.org>
While most display helpers Kconfig symbols have the DRM_DISPLAY prefix,
the DisplayPort-AUX chardev interface uses DRM_DP_AUX_CHARDEV.
Since the number of users is limited and it's a selected symbol, we can
easily rename it to make it consistent.
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Reviewed-by: Lucas De Marchi <lucas.demarchi@intel.com>
Link: https://lore.kernel.org/r/20240327-kms-kconfig-helpers-v3-3-eafee11b84b3@kernel.org
Signed-off-by: Maxime Ripard <mripard@kernel.org>
While most display helpers Kconfig symbols have the DRM_DISPLAY prefix,
the DisplayPort Tunnel debugging uses DRM_DISPLAY_DEBUG_DP_TUNNEL_STATE.
Since the number of users is limited and it's a selected symbol, we can
easily rename it to make it consistent.
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Reviewed-by: Lucas De Marchi <lucas.demarchi@intel.com>
Link: https://lore.kernel.org/r/20240327-kms-kconfig-helpers-v3-2-eafee11b84b3@kernel.org
Signed-off-by: Maxime Ripard <mripard@kernel.org>
While most display helpers Kconfig symbols have the DRM_DISPLAY prefix,
the DisplayPort AUX bus implementation uses DRM_DP_AUX_BUS.
Since the number of users is limited and it's a selected symbol, we can
easily rename it to make it consistent.
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Reviewed-by: Lucas De Marchi <lucas.demarchi@intel.com>
Link: https://lore.kernel.org/r/20240327-kms-kconfig-helpers-v3-1-eafee11b84b3@kernel.org
Signed-off-by: Maxime Ripard <mripard@kernel.org>
The init sequence specifies the 0x11 and 0x29 dsi commands, which are
the exit-sleep and display-on commands.
In the actual prepare step the driver already uses the appropriate
function calls for those, so drop the duplicates.
Fixes: e5f9d543419c ("drm/panel: ltk050h3146w: add support for Leadtek LTK050H3148W-CTA6 variant")
Signed-off-by: Heiko Stuebner <heiko.stuebner@cherry.de>
Reviewed-by: Quentin Schulz <quentin.schulz@theobroma-systems.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240320131232.327196-2-heiko@sntech.de
Similar to other variants, the LTK050H3148W wants to run in video mode
when displaying data. So far only the Synopsis DSI driver was using this
panel and it is always switching to video mode, independent of this flag
being set.
Other DSI drivers might handle this differently, so add the flag.
Fixes: e5f9d543419c ("drm/panel: ltk050h3146w: add support for Leadtek LTK050H3148W-CTA6 variant")
Signed-off-by: Heiko Stuebner <heiko.stuebner@cherry.de>
Reviewed-by: Quentin Schulz <quentin.schulz@theobroma-systems.com>
Acked-by: Jessica Zhang <quic_jesszhan@quicinc.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240320131232.327196-1-heiko@sntech.de
If job accounting is disabled, then both fdinfo's drm-engine and drm-cycle
key values will remain immutable. In that case, it makes more sense not to
display them at all to avoid confusing user space profiling tools.
Signed-off-by: Adrián Larumbe <adrian.larumbe@collabora.com>
Reviewed-by: Steven Price <steven.price@arm.com>
Signed-off-by: Steven Price <steven.price@arm.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240316231306.293817-1-adrian.larumbe@collabora.com
The array size calculation in pvr_vm_mips_fini() appears to be incorrect based on
taking the size of the pointer rather than the size of the array, which manifests
as a warning about signed integer overflow:
In file included from include/linux/kernel.h:16,
from drivers/gpu/drm/imagination/pvr_rogue_fwif.h:10,
from drivers/gpu/drm/imagination/pvr_ccb.h:7,
from drivers/gpu/drm/imagination/pvr_device.h:7,
from drivers/gpu/drm/imagination/pvr_vm_mips.c:4:
drivers/gpu/drm/imagination/pvr_vm_mips.c: In function 'pvr_vm_mips_fini':
include/linux/array_size.h:11:25: error: overflow in conversion from 'long unsigned int' to 'int' changes value from '18446744073709551615' to '-1' [-Werror=overflow]
11 | #define ARRAY_SIZE(arr) (sizeof(arr) / sizeof((arr)[0]) + __must_be_array(arr))
| ^
drivers/gpu/drm/imagination/pvr_vm_mips.c:106:24: note: in expansion of macro 'ARRAY_SIZE'
106 | for (page_nr = ARRAY_SIZE(mips_data->pt_pages) - 1; page_nr >= 0; page_nr--) {
| ^~~~~~~~~~
Just use the number of array elements directly here, and in the corresponding
init function for consistency.
Fixes: 927f3e0253c1 ("drm/imagination: Implement MIPS firmware processor and MMU support")
Reviewed-by: Donald Robson <donald.robson@imgtec.com>
Link: https://lore.kernel.org/lkml/9df9e4f87727399928c068dbbf614c9895ae15f9.camel@imgtec.com/
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Matt Coster <matt.coster@imgtec.com>
virt_to_pfn() isn't available on x86 (except to xen) so breaks
COMPILE_TEST builds. Avoid its use completely by instead storing the
struct page pointer allocated in panthor_device_init() and using
page_to_pfn() instead.
Signed-off-by: Steven Price <steven.price@arm.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240318145119.368582-1-steven.price@arm.com
Putting a hard dependency on CONFIG_PM is not possible because of a
circular dependency issue, and it's actually not desirable either. In
order to support this use case, we forcibly resume at init time, and
suspend at unplug time.
v2:
- Drop the #ifdef CONFIG_PM section around panthor_pm_ops's definition
Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202403031944.EOimQ8WK-lkp@intel.com/
Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Steven Price <steven.price@arm.com>
Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240318153117.1321544-1-boris.brezillon@collabora.com
When the atna33xc20 driver was first written the resume code never
returned an error. If there was a problem waiting for HPD it just
printed a warning and moved on. This changed in response to review
feedback [1] on a future patch but I accidentally didn't account for
rolling back the regulator enable in the error cases. Do so now.
[1] https://lore.kernel.org/all/5f3cf3a6-1cc2-63e4-f76b-4ee686764705@linaro.org/
Fixes: 3b5765df375c ("drm/panel: atna33xc20: Take advantage of wait_hpd_asserted() in struct drm_dp_aux")
Acked-by: Jessica Zhang <quic_jesszhan@quicinc.com>
Signed-off-by: Douglas Anderson <dianders@chromium.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20240313-homestarpanel-regulator-v1-1-b8e3a336da12@chromium.org
The BOs in a bulk move must share all the same reservation object
to make sure that we lock the whole bulk during eviction.
Actually document and enforce that with a warning.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240312105555.3065-1-christian.koenig@amd.com
In case the LCDIF is enabled in DT but unused, the clocks used by the
LCDIF are not enabled. Those clocks may even have a use count of 0 in
case there are no other users of those clocks. This can happen e.g. in
case the LCDIF drives HDMI bridge which has no panel plugged into the
HDMI connector.
Do not attempt to disable clocks in the suspend callback and re-enable
clocks in the resume callback unless the LCDIF is enabled and was in
use before the system entered suspend, otherwise the driver might end
up trying to disable clocks which are already disabled with use count
0, and would trigger a warning from clock core about this condition.
Note that the lcdif_rpm_suspend() and lcdif_rpm_resume() functions
internally perform the clocks disable and enable operations and act
as runtime PM hooks too.
Reviewed-by: Liu Ying <victor.liu@nxp.com>
Fixes: 9db35bb349a0 ("drm: lcdif: Add support for i.MX8MP LCDIF variant")
Signed-off-by: Marek Vasut <marex@denx.de>
Link: https://patchwork.freedesktop.org/patch/msgid/20240226082644.32603-1-marex@denx.de
Commit 00084f0c01bf ("drm: bridge: thc63lvd1024: Switch to use
of_graph_get_remote_node()") simplified the thc63lvd1024 driver by
replacing hand-rolled code with a helper function. While doing so, it
created an error code path at probe time without any error message,
potentially causing probe issues that get annoying to debug. Fix it by
adding an error message.
Fixes: 00084f0c01bf ("drm: bridge: thc63lvd1024: Switch to use of_graph_get_remote_node()")
Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Reviewed-by: Neil Armstrong <neil.armstrong@linaro.org>
Link: https://lore.kernel.org/r/20240318160601.2813-1-laurent.pinchart+renesas@ideasonboard.com
Signed-off-by: Neil Armstrong <neil.armstrong@linaro.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20240318160601.2813-1-laurent.pinchart+renesas@ideasonboard.com
If a specific design doesn't wire IT66121's interrupt signal output pin up
to the display controller side, then we should not register the interrupt
handler. Such a decision is valid usage, as we can fall back to polling
mode. So, don't make the assumption that a specific board always supports
HPD. Carry out a sanity check on 'client->irq' before using it, fall back
to polling mode if client->irq < 0 is true. Such a design increases the
overall flexibility.
Signed-off-by: Sui Jingfeng <sui.jingfeng@linux.dev>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://lore.kernel.org/r/20240316160536.1051513-1-sui.jingfeng@linux.dev
Signed-off-by: Neil Armstrong <neil.armstrong@linaro.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20240316160536.1051513-1-sui.jingfeng@linux.dev
Commit 95da53d63dcf ("drm/omapdrm: Use regular fbdev I/O helpers")
stopped console from updating for command mode displays because there is
no damage handling in fb_sys_write() unlike we had earlier in
drm_fb_helper_sys_write().
Let's fix the issue by adding FB_GEN_DEFAULT_DEFERRED_DMAMEM_OPS and
FB_DMAMEM_HELPERS_DEFERRED as suggested by Thomas. We cannot use the
FB_DEFAULT_DEFERRED_OPS as fb_deferred_io_mmap() won't work properly
for write-combine.
Fixes: 95da53d63dcf ("drm/omapdrm: Use regular fbdev I/O helpers")
Suggested-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Thomas Zimmermann <tzimmermann@suse.de>
Signed-off-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240228063540.4444-3-tony@atomide.com
The framebuffer console stopped updating with commit f231af498c29
("drm/fb-helper: Disconnect damage worker from update logic").
Let's fix the issue by implementing fb_dirty similar to what was done
with commit 039a72ce7e57 ("drm/i915/fbdev: Implement fb_dirty for intel
custom fb helper").
Fixes: f231af498c29 ("drm/fb-helper: Disconnect damage worker from update logic")
Reviewed-by: Thomas Zimmermann <tzimmermann@suse.de>
Signed-off-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240228063540.4444-2-tony@atomide.com
I've recently been seeing some unexplained GSP errors on my RTX 6000 from
failed aux transactions:
[ 132.915867] nouveau 0000:1f:00.0: gsp: cli:0xc1d00002 obj:0x00730000
ctrl cmd:0x00731341 failed: 0x0000ffff
While the cause of these is not yet clear, these messages made me notice
that the aux transactions causing these transactions were succeeding - not
failing. As it turns out, this is because we're currently not returning the
correct variable when r535_dp_aux_xfer() hits an error - causing us to
never propagate GSP errors for failed aux transactions to userspace.
So, let's fix that.
Fixes: 4ae3a20102b2 ("nouveau/gsp: don't free ctrl messages on errors")
Signed-off-by: Lyude Paul <lyude@redhat.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240315212104.776936-1-lyude@redhat.com
There are 2 different AUO panels using the same panel id. One of the
variants requires using overridden modes to resolve glitching issue as
described in commit 70e0d5550f5c ("drm/panel-edp: Add auo_b116xa3_mode").
Other variants should use the modes parsed from EDID.
Signed-off-by: Hsin-Yi Wang <hsinyi@chromium.org>
Reviewed-by: Douglas Anderson <dianders@chromium.org>
Signed-off-by: Douglas Anderson <dianders@chromium.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20240307230653.1807557-6-hsinyi@chromium.org
It's found that some panels have variants that they share the same panel id
although their EDID and names are different. When matching generic edp
panels, we should first match with both panel identity, which contains both
panel id and panel name. If not found, match with panel id only.
Signed-off-by: Hsin-Yi Wang <hsinyi@chromium.org>
Reviewed-by: Douglas Anderson <dianders@chromium.org>
Signed-off-by: Douglas Anderson <dianders@chromium.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20240307230653.1807557-5-hsinyi@chromium.org
Currently edid quirks are matched by panel id only.
Modify it to match with identity so it's easier to be extended
for more complex matching if required.
Signed-off-by: Hsin-Yi Wang <hsinyi@chromium.org>
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Reviewed-by: Douglas Anderson <dianders@chromium.org>
Signed-off-by: Douglas Anderson <dianders@chromium.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20240307230653.1807557-4-hsinyi@chromium.org
Create a type drm_edid_ident as the identity of an EDID. Currently it
contains panel id and monitor name.
Create a function that can match a given EDID and an identity:
1. Reject if the panel id doesn't match.
2. If name is not null in identity, try to match it in the detailed timing
blocks. Note that some panel vendors put the monitor name after
EDID_DETAIL_MONITOR_STRING.
Signed-off-by: Hsin-Yi Wang <hsinyi@chromium.org>
Reviewed-by: Douglas Anderson <dianders@chromium.org>
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Signed-off-by: Douglas Anderson <dianders@chromium.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20240307230653.1807557-3-hsinyi@chromium.org
It's found that some panels have variants that they share the same panel id
although their EDID and names are different. Besides panel id, now we need
more information from the EDID base block to distinguish these panel
variants.
Add drm_edid_read_base_block() to return the EDID base block, which is
wrapped in struct drm_edid.
Caller can further use it to get panel id or check if the block contains
certain strings, such as panel name.
Merge drm_edid_get_panel_id() and edid_extract_panel_id() into one
function.
Signed-off-by: Hsin-Yi Wang <hsinyi@chromium.org>
Reviewed-by: Douglas Anderson <dianders@chromium.org>
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Signed-off-by: Douglas Anderson <dianders@chromium.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20240307230653.1807557-2-hsinyi@chromium.org
Pin and vmap are distinct operations. Do not perform a pin as part
of the vmap call. This used to be necessary to keep the fbdev buffer
in place while it is being updated. Fbdev emulation has meanwhile
been fixed to lock the buffer correctly. Same for vunmap.
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Dmitry Osipenko <dmitry.osipenko@collabora.com>
Tested-by: Dmitry Osipenko <dmitry.osipenko@collabora.com> # virtio-gpu
Acked-by: Christian König <christian.koenig@amd.com>
Acked-by: Zack Rusin <zack.rusin@broadcom.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240227113853.8464-14-tzimmermann@suse.de
Pin and vmap are distinct operations. Do not perform a pin as part
of the vmap call. This used to be necessary to keep the fbdev buffer
in place while it is being updated. Fbdev emulation has meanwhile
been fixed to lock the buffer correctly. Same for vunmap.
For refactoring the code, remove the pin calls from the helper's
vmap implementation in drm_gem_vram_vmap() and inline the call to
drm_gem_vram_kmap_locked(). This gives a vmap helper that only
maps the buffer object's memory pages without pinning or locking.
Do a similar refactoring for vunmap.
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Dmitry Osipenko <dmitry.osipenko@collabora.com>
Tested-by: Dmitry Osipenko <dmitry.osipenko@collabora.com> # virtio-gpu
Acked-by: Zack Rusin <zack.rusin@broadcom.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240227113853.8464-13-tzimmermann@suse.de
The function drm_client_buffer_vmap() establishes a long-term mapping
of the client's buffer object into the kernel address space. Make sure
that buffer does not move by pinning it to its current location. Same
for vunmap with unpin.
The only caller of drm_client_buffer_vmap() is fbdev-dma, which uses
gem-dma. As DMA-backed GEM buffers do not move, this change is for
correctness with little impact in practice.
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Dmitry Osipenko <dmitry.osipenko@collabora.com>
Tested-by: Dmitry Osipenko <dmitry.osipenko@collabora.com> # virtio-gpu
Acked-by: Zack Rusin <zack.rusin@broadcom.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240227113853.8464-12-tzimmermann@suse.de
Temporarily lock the fbdev buffer object during updates to prevent
memory managers from evicting/moving the buffer. Moving a buffer
object while update its content results in undefined behaviour.
Fbdev-generic updates its buffer object from a shadow buffer. Gem-shmem
and gem-dma helpers do not move buffer objects, so they are safe to be
used with fbdev-generic. Gem-vram and qxl are based on TTM, but pin
buffer objects are part of the vmap operation. So both are also safe
to be used with fbdev-generic.
Amdgpu and nouveau do not pin or lock the buffer object during an
update. Their TTM-based memory management could move the buffer object
while the update is ongoing.
The new vmap_local and vunmap_local helpers hold the buffer object's
reservation lock during the buffer update. This prevents moving the
buffer object on all memory managers.
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Dmitry Osipenko <dmitry.osipenko@collabora.com>
Tested-by: Dmitry Osipenko <dmitry.osipenko@collabora.com> # virtio-gpu
Acked-by: Zack Rusin <zack.rusin@broadcom.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240227113853.8464-11-tzimmermann@suse.de
Acquire the buffer object's reservation lock in drm_gem_pin() and
remove locking the drivers' GEM callbacks where necessary. Same for
unpin().
DRM drivers and memory managers modified by this patch will now have
correct dma-buf locking semantics: the caller is responsible for
holding the reservation lock when calling the pin or unpin callback.
DRM drivers and memory managers that are not modified will now be
protected against concurent invocation of their pin and unpin callbacks.
PRIME does not implement struct dma_buf_ops.pin, which requires
the caller to hold the reservation lock. It does implement struct
dma_buf_ops.attach, which requires to callee to acquire the
reservation lock. The PRIME code uses drm_gem_pin(), so locks
are now taken as specified. Same for unpin and detach.
The patch harmonizes GEM pin and unpin to have non-interruptible
reservation locking across all drivers, as is already the case for
vmap and vunmap. This affects gem-shmem, gem-vram, loongson, qxl and
radeon.
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Zack Rusin <zack.rusin@broadcom.com>
Reviewed-by: Dmitry Osipenko <dmitry.osipenko@collabora.com>
Tested-by: Dmitry Osipenko <dmitry.osipenko@collabora.com> # virtio-gpu
Link: https://patchwork.freedesktop.org/patch/msgid/20240227113853.8464-10-tzimmermann@suse.de
Acquire the reservation lock directly in GEM pin callback. Same for
unpin. Prepares for further changes.
Dma-buf locking semantics require callers to hold the buffer's
reservation lock when invoking the pin and unpin callbacks. Prepare
qxl accordingly by pushing locking out of the implementation. A
follow-up patch will fix locking for all GEM code at once.
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Dmitry Osipenko <dmitry.osipenko@collabora.com>
Tested-by: Dmitry Osipenko <dmitry.osipenko@collabora.com> # virtio-gpu
Acked-by: Christian König <christian.koenig@amd.com>
Acked-by: Zack Rusin <zack.rusin@broadcom.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240227113853.8464-9-tzimmermann@suse.de
Rename __qxl_bo_pin() to qxl_bo_pin_locked() and update all callers.
The function will be helpful for implementing the GEM pin callback
with correct semantics. Same for __qxl_bo_unpin().
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Dmitry Osipenko <dmitry.osipenko@collabora.com>
Tested-by: Dmitry Osipenko <dmitry.osipenko@collabora.com> # virtio-gpu
Acked-by: Christian König <christian.koenig@amd.com>
Acked-by: Zack Rusin <zack.rusin@broadcom.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240227113853.8464-8-tzimmermann@suse.de
Acquire the reservation lock directly in GEM pin callback. Same for
unpin. Prepares for further changes.
Dma-buf locking semantics require callers to hold the buffer's
reservation lock when invoking the pin and unpin callbacks. Prepare
nouveau accordingly by pushing locking out of the implementation. A
follow-up patch will fix locking for all GEM code at once.
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Dmitry Osipenko <dmitry.osipenko@collabora.com>
Tested-by: Dmitry Osipenko <dmitry.osipenko@collabora.com> # virtio-gpu
Acked-by: Christian König <christian.koenig@amd.com>
Acked-by: Zack Rusin <zack.rusin@broadcom.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240227113853.8464-7-tzimmermann@suse.de
