Daniel Vetter [Tue, 27 Aug 2024 10:55:16 +0000 (12:55 +0200)]
Merge tag 'drm-intel-gt-next-2024-08-23' of https://gitlab.freedesktop.org/drm/i915/kernel into drm-next
UAPI Changes:
- Limit the number of relocations to INT_MAX (Tvrtko)
Only impact should be synthetic tests.
Driver Changes:
- Fix for #11396: GPU Hang and rcs0 reset on Cherrytrail platform
- Fix Virtual Memory mapping boundaries calculation (Andi)
- Fix for #11255: Long hangs in buddy allocator with DG2/A380 without
Resizable BAR since 6.9 (David)
- Mark the GT as dead when mmio is unreliable (Chris, Andi)
- Workaround additions / fixes for MTL, ARL and DG2 (John H, Nitin)
- Enable partial memory mapping of GPU virtual memory (Andi, Chris)
- Prevent NULL deref on intel_memory_regions_hw_probe (Jonathan, Dan)
- Avoid UAF on intel_engines_release (Krzysztof)
Stefan Wahren [Wed, 21 Aug 2024 21:40:46 +0000 (23:40 +0200)]
drm/vc4: Get the rid of DRM_ERROR()
DRM_ERROR() has been deprecated in favor of pr_err(). However, we
should prefer to use drm_err() whenever possible so we get device-
specific output with the error message. In error case of kcalloc,
we can simply drop DRM_ERROR(), because kcalloc already logs errors.
Stefan Wahren [Wed, 21 Aug 2024 21:40:45 +0000 (23:40 +0200)]
drm/vc4: hdmi: Handle error case of pm_runtime_resume_and_get
The commit 0f5251339eda ("drm/vc4: hdmi: Make sure the controller is
powered in detect") introduced the necessary power management handling
to avoid register access while controller is powered down.
Unfortunately it just print a warning if pm_runtime_resume_and_get()
fails and proceed anyway.
This could happen during suspend to idle. So we must assume it is unsafe
to access the HDMI register. So bail out properly.
Ast's BMC connector tracks the status of an underlying physical
connector and updates the BMC status accordingly. This functionality
works around GNOME's settings app, which cannot handle multiple
outputs on the same CRTC.
The workaround is now obsolete as all code for physical outputs
handle BMC support internally. Hence, remove the driver's code and
the BMC output entirely.
Permanently set the connector status to 'connected'. Return BMC modes
for connector if no display is attached to the physical DP connector.
Otherwise use EDID modes as before.
If the status of the physical connector changes, the driver still
generates a hotplug event. DRM clients will then reconfigure their
output to a mode appropriate for either physical display or BMC.
v3:
- use struct ast_connector.physical_status to handle BMC
Permanently set the connector status to 'connected'. Return BMC modes
for connector if no display is attached to the physical DP connector.
Otherwise use EDID modes as before.
If the status of the physical connector changes, the driver still
generates a hotplug event. DRM clients will then reconfigure their
output to a mode appropriate for either physical display or BMC.
v3:
- use struct ast_connector.physical_status to handle BMC
Permanently set the connector status to 'connected'. Return BMC modes
for connector if no display is attached to the physical DP connector.
Otherwise use EDID modes as before.
If the status of the physical connector changes, the driver still
generates a hotplug event. DRM clients will then reconfigure their
output to a mode appropriate for either physical display or BMC.
v3:
- use struct ast_connector.physical_status to handle BMC
Convert DP501 support to struct drm_edid and its helpers. Simplifies
and modernizes the EDID handling.
The driver reads 4 bytes at once, but the overall read length is now
variable. Therefore update the EDID read loop to never return more than
the requested bytes.
Permanently set the connector status to 'connected'. Return BMC modes
for connector if no display is attached to the physical DP connector.
Otherwise use EDID modes as before.
If the status of the physical connector changes, the driver still
generates a hotplug event. DRM clients will then reconfigure their
output to a mode appropriate for either physical display or BMC.
v3:
- use struct ast_connector.physical_status to handle BMC
drm/ast: astdp: Simplify power management when detecting display
Remove the CRTC handling in the ASTDP detect_ctx helper and enable
power while the detecting the display. Unconditionally wait a few
milliseconds after switching power. Simplifies the code and makes it
more robust.
Convert ASTDP support to struct drm_edid and its helpers. Simplifies
and modernizes the EDID handling.
The driver reads 4 bytes at once, but the overall read length is now
variable. Therefore update the EDID read loop to never return more than
the requested bytes.
The device does not seem to support EDID extensions, as the driver
actively clears any such information from the main EDID header. As
the new interface allows for reading extension blocks for EDID, make
sure that the block is always 0 (i.e., the main header). A later
update might fix that.
The modeset mutex protects EDID retrival from concurrent modeset
operations. Acquire the lock in ast_astdp_read_edid(). Prepares the
code for conversion to struct drm_edid.
Add struct ast_connector to track a connector's physical status. With
the upcoming BMC support, the physical status can be different from the
reported status.
Andi Shyti [Wed, 7 Aug 2024 10:05:21 +0000 (11:05 +0100)]
drm/i915/gem: Calculate object page offset for partial memory mapping
To enable partial memory mapping of GPU virtual memory, it's
necessary to introduce an offset to the object's memory
(obj->mm.pages) scatterlist. This adjustment compensates for
instances when userspace mappings do not start from the beginning
of the object.
Andi Shyti [Wed, 7 Aug 2024 10:05:20 +0000 (11:05 +0100)]
drm/i915/gem: Do not look for the exact address in node
In preparation for the upcoming partial memory mapping feature,
we want to make sure that when looking for a node we consider
also the offset and not just the starting address of the virtual
memory node.
mipi_dsi_dcs_set_tear_scanline_multi can heavily benefit from being
converted to a multi style function as it is often called in the context of
similar functions.
The Mediatek Soc DSI host has different modes in prepare() and
enable() functions, prepare() is in LP mode and enable() is in
HS mode. Since the "exit sleep mode" and "set display on"
command must also be sent in LP mode, so we also move "exit
sleep mode" and "set display on" command to the init() function.
We have no other actions in the enable() function after moves
"exit sleep mode" and "set display on", and we checked the call
of the enable() function during the "startup" process. It seems
that only one judgment was made in drm_panel_enabel(). If the
panel does not define enable(), the judgment will skip the
enable() and continue execution. This does not seem to have
any other effect, and we found that some drivers also seem
to have no enable() function added, for example:
panel-asus-z00t-tm5p5-n35596 / panel-boe-himax8279d...
In addition, we briefly tested the kingdisplay_kd101ne3 panel and
melfas_lmfbx101117480 panel, and it seems that there is no garbage
on the panel, so we delete enable() function.
After moving the "exit sleep mode" and "set display on" command
to the init() function, we no longer need additional delay
judgment, so we delete variables "exit_sleep_to_display_on_delay_ms"
and "display_on_delay_ms".
drm/panel: st7701: Add Anbernic RG28XX panel support
The Anbernic RG28XX is a handheld gaming device with a 2.8 inch 480x640
display. Add support for the display panel.
This panel is driven by a variant of ST7701 driver IC internally,
confirmed by dumping and analyzing its BSP initialization sequence
by using a logic analyzer. It is very similar to the existing
densitron,dmt028vghmcmi-1a panel, but differs in some unknown
register values. Besides, it is connected via SPI, so add a new entry
for the panel.
drm/panel: st7701: Add support for SPI for configuration
The ST7701 supports not only MIPI DSI, but also SPI as an interface
for configuration. To support a panel connected via SPI with an RGB
parallel interface, add support for SPI using MIPI DBI helpers.
The RG28XX panel is a display panel of the Anbernic RG28XX, a handheld
gaming device from Anbernic. It is 2.8 inches in size (diagonally) with
a resolution of 480x640.
This panel is driven by a variant of the ST7701 driver IC internally,
confirmed by dumping and analyzing its BSP initialization sequence
by using a logic analyzer. It is very similar to the existing
densitron,dmt028vghmcmi-1a panel, but differs in some unknown
register values, so add a new entry for the panel to distinguish them.
Additionally, the panel only has an SPI instead of MIPI DSI.
So add and modify for SPI as well.
Although a panel connected via SPI with a D/CX pin theoretically exists,
no such panels have been found for this driver yet. Therefore, disable
the use of the dc-gpios property for all currently known devices.
Esben Haabendal [Wed, 14 Aug 2024 10:37:26 +0000 (12:37 +0200)]
drm/bridge: nwl-dsi: Use vsync/hsync polarity from display mode
Using the correct bit helps. The documentation specifies bit 0 in both
registers to be controlling polarity of dpi_vsync_input and
dpi_hsync_input polarity. Bit 1 is reserved, and should therefore not be
set.
Tested with panel that requires active high vsync and hsync.
drm: bridge: anx7625: Use of_property_read_variable_u8_array()
There's no need to get the length of an DT array property before
parsing the array. of_property_read_variable_u8_array() takes a minimum
and maximum length and returns the actual length (or error code).
This is part of a larger effort to remove callers of of_get_property()
and similar functions. of_get_property() leaks the DT property data
pointer which is a problem for dynamically allocated nodes which may
be freed.
Make use of devm_clk_get_enabled() to replace devm_clk_get() and
clk_prepare_enable() for isfr and iahb clocks, and drop the now
unnecessary calls to clk_disable_unprepare().
Similarly, use devm_clk_get_optional_enabled() helper for cec clock,
which additionally allows to remove the -ENOENT test.
Moreover, the clock related members of struct dw_hdmi are not required
anymore, hence drop them.
Liu Ying [Tue, 13 Aug 2024 09:16:37 +0000 (17:16 +0800)]
drm/bridge: lontium-lt8912b: Validate mode in drm_bridge_funcs::mode_valid()
If the bridge is attached with the DRM_BRIDGE_ATTACH_NO_CONNECTOR flag set,
this driver won't initialize a connector and hence display mode won't be
validated in drm_connector_helper_funcs::mode_valid(). So, move the mode
validation from drm_connector_helper_funcs::mode_valid() to
drm_bridge_funcs::mode_valid(), because the mode validation is always done
for the bridge.
Luca Coelho [Tue, 13 Aug 2024 14:06:18 +0000 (17:06 +0300)]
drm/i915/gt: remove stray declaration of intel_gt_release_all()
When intel_gt_release_all() was removed from the code in commit e89950553385 ("drm/i915: do not clean GT table on error path"), its
declaration in the header file remained. Remove it.
Jani Nikula [Wed, 14 Aug 2024 10:00:34 +0000 (13:00 +0300)]
string: add mem_is_zero() helper to check if memory area is all zeros
Almost two thirds of the memchr_inv() usages check if the memory area is
all zeros, with no interest in where in the buffer the first non-zero
byte is located. Checking for !memchr_inv(s, 0, n) is also not very
intuitive or discoverable. Add an explicit mem_is_zero() helper for this
use case.
Replace FB_BLANK_ constants with their counterparts from the
backlight subsystem. The values are identical, so there's no
change in functionality or semantics.
Replace FB_BLANK_ constants with their counterparts from the
backlight subsystem. The values are identical, so there's no
change in functionality or semantics.
drm/panel: panel-sony-acx565akm: Use backlight power constants
Replace FB_BLANK_ constants with their counterparts from the
backlight subsystem. The values are identical, so there's no
change in functionality or semantics.
drm/panel: panel-samsung-s6e3ha2: Use backlight power constants
Replace FB_BLANK_ constants with their counterparts from the
backlight subsystem. The values are identical, so there's no
change in functionality or semantics.
drm/panel: panel-samsung-s6e63j0x03: Use backlight power constants
Replace FB_BLANK_ constants with their counterparts from the
backlight subsystem. The values are identical, so there's no
change in functionality or semantics.
drm/panel: panel-orisetech-otm8009a: Use backlight power constants
Replace FB_BLANK_ constants with their counterparts from the
backlight subsystem. The values are identical, so there's no
change in functionality or semantics.
drm/panel: panel-novatak-nt35510: Use backlight power constants
Replace FB_BLANK_ constants with their counterparts from the
backlight subsystem. The values are identical, so there's no
change in functionality or semantics.
Replace FB_BLANK_ constants with their counterparts from the
backlight subsystem. The values are identical, so there's no
change in functionality or semantics.
Add max_tmds_clock validation to prepare for additions and changes to
the MPLL config table. Use the same rate restrictions that is currently
applied.
The rate limit for RK3288, RK3399 and RK3568 is based on current mpll
table. The rate limit for RK3228 and RK3328 is based on the
inno-hdmi-phy pre-pll table.
drm/rockchip: dw_hdmi: Drop superfluous assignments of mpll_cfg, cur_ctr and phy_config
The mpll_cfg, cur_ctr and phy_config members in struct dw_hdmi_plat_data
are only used to configure the Synopsys PHYs supported internally by DW
HDMI transmitter driver (gpu/drm/bridge/synopsys/dw-hdmi.c), via
hdmi_phy_configure_dwc_hdmi_3d_tx(), which is further invoked from
dw_hdmi_phy_init(). This is part of the internal
dw_hdmi_synopsys_phy_ops struct, managed within dw_hdmi_detect_phy().
To handle vendor PHYs, DW HDMI driver doesn't make use of the internal
PHY ops and, instead, relies on the glue layer to provide the phy_ops
and phy_name members of struct dw_hdmi_plat_data.
Drop the unnecessary assignments of DW internal PHY related members from
structs rk3228_hdmi_drv_data and rk3328_hdmi_drv_data, since both set
the phy_force_vendor flag and correctly provide the expected vendor PHY
data.
drm/rockchip: dw_hdmi: Use devm_regulator_get_enable()
The regulators are only enabled at bind() and disabled at unbind(),
hence replace the boilerplate code by making use of
devm_regulator_get_enable() helper.
Make use of devm_clk_get_optional_enabled() to replace devm_clk_get()
and clk_prepare_enable() for ref_clk and drop the now unnecessary calls
to clk_disable_unprepare().
Additionally, use devm_clk_get_optional() helper for grf_clk to replace
the open coding call to devm_clk_get() followed by the -ENOENT test.
Val Packett [Mon, 24 Jun 2024 20:40:48 +0000 (17:40 -0300)]
drm/rockchip: vop: clear DMA stop bit on RK3066
The RK3066 VOP sets a dma_stop bit when it's done scanning out a frame
and needs the driver to acknowledge that by clearing the bit.
Unless we clear it "between" frames, the RGB output only shows noise
instead of the picture. atomic_flush is the place for it that least
affects other code (doing it on vblank would require converting all
other usages of the reg_lock to spin_(un)lock_irq, which would affect
performance for everyone).
This seems to be a redundant synchronization mechanism that was removed
in later iterations of the VOP hardware block.
drm/rockchip: cdn-dp: Clean up a few logged messages
Clean up a few logged messages, which were previously worded as rather
incomplete sentences separated by periods. This was both a bit unreadable
and grammatically incorrect, so convert them into partial sentences separated
(or connected) by semicolons, together with some wording improvements.
Jani Nikula [Mon, 12 Aug 2024 14:28:48 +0000 (17:28 +0300)]
drm/i915/gvt: stop using drm_edid_block_valid()
We'll want to stop drm_edid_block_valid() usage. KVMGT is the last
user. Replace with drm_edid_valid(), which unfortunately requires an
allocated drm_edid. However, on the plus side, this would be required to
handle the TODO comment about EDID extension block support.
Jani Nikula [Tue, 14 May 2024 12:55:07 +0000 (15:55 +0300)]
drm/rockchip: cdn-dp: get rid of drm_edid_raw()
The dimensions are available in display info, so there's no need for raw
EDID access. While at it, move the debug logging to where the EDID is
actually read.
Jani Nikula [Thu, 6 Jun 2024 12:35:03 +0000 (15:35 +0300)]
drm/edid: reduce DisplayID log spamming
Debug printing at DisplayID validation leads to lots of log spamming as
it's called at DisplayID iterators during EDID parsing. Remove it, and
replace with a less noisy message at connector EDID update.
A vulnerability in the package_index module of pypa/setuptools
versions up to 69.1.1 allows for remote code execution via its
download functions. These functions, which are used to download
packages from URLs provided by users or retrieved from package
index servers, are susceptible to code injection. If these
functions are exposed to user-controlled inputs, such as package
URLs, they can execute arbitrary commands on the system. The
issue is fixed in version 70.0.
Severity: 8.8 / 10 (High)
Attack vector: Network
Attack complexity: Low
Privileges required: None
User interaction: Required
Scope: Unchanged
Confidentiality: High
Integrity: High
Availability: High
CVE ID: CVE-2024-6345"
To avoid disturbing everyone with the kernel repo hosted on GitHub,
I suggest we upgrade our python dependencies once again to appease
GitHub Dependabot.
Tejas Vipin [Tue, 6 Aug 2024 13:59:48 +0000 (19:29 +0530)]
drm/mipi-dsi: add more multi functions for better error handling
Add more functions that can benefit from being multi style and mark
older variants as deprecated to eventually convert all mipi_dsi functions
to multi style.
Dan Carpenter [Mon, 12 Aug 2024 08:29:38 +0000 (11:29 +0300)]
drm/ast: astdp: fix loop timeout check
This code has an issue because it loops until "i" is set to UINT_MAX but
the test for failure assumes that "i" is set to zero. The result is that
it will only print an error message if we succeed on the very last try.
Reformat the loop to count forwards instead of backwards.
Zhang Zekun [Mon, 12 Aug 2024 12:35:43 +0000 (20:35 +0800)]
drm/hisilicon: Remove unused delarations
hibmc_mm_init() has been removed since commit 28645ae064d1
("drm/hisilicon/hibmc: Remove hibmc_ttm.c"), but remain the declaration
untouched in the header files. So, let's remove this unused declaration.
The output_poll_changed hook in struct drm_mode_config_funcs is
unused. Remove it. The helper drm_client_dev_hotplug() implements
the callback's functionality.
drm/nouveau: Implement switcheroo reprobe with drm_client_dev_hotplug()
Replace the call to drm_fb_helper_output_poll_changed() with a call
to drm_client_dev_hotplug(). It is equivalent in functionality, but
uses the DRM client infrastructure.
drm/nouveau: Do not set struct drm_mode_config_funcs.output_poll_changed
The output_poll_changed hook was only necessary before in-kernel
DRM clients existed, but is now obsolete. The client code handles
display hotplugging internally.
drm/nouveau: Do not set struct drm_driver.lastclose
Remove the implementation of struct drm_driver.lastclose. The hook
was only necessary before in-kernel DRM clients existed, but is now
obsolete. The code in nouveau_vga_lastclose() is performed by
drm_lastclose().
drm/amdgpu: Do not set struct drm_driver.lastclose
Remove the implementation of struct drm_driver.lastclose. The hook
was only necessary before in-kernel DRM clients existed, but is now
obsolete. The code in amdgpu_driver_lastclose_kms() is performed by
drm_lastclose().
Amdgpu and nouveau call vga_switcheroo_process_delayed_switch() from
their lastclose callbacks. Call it from drm_lastclose(), so that the
driver functions can finally be removed. Only PCI devices with enabled
switcheroo do the delayed switching. The call has no effect on other
hardware.
v2:
- move change to drm_lastclose() (Sima)
- update docs for vga_switcheroo_process_delayed_switch()
Daniel Yang [Fri, 9 Aug 2024 03:23:50 +0000 (20:23 -0700)]
drm/connector: kerneldoc: Fix two missing newlines in drm_connector.c
Fix the unexpected indentation errors.
drm_connector.c has some kerneldoc comments that were missing newlines.
This results in the following warnings when running make htmldocs:
./Documentation/gpu/drm-kms:538: ./drivers/gpu/drm/drm_connector.c:2344: WARNING: Definition list ends without a blank line; unexpected unindent. [docutils]
./Documentation/gpu/drm-kms:538: ./drivers/gpu/drm/drm_connector.c:2346: ERROR: Unexpected indentation. [docutils]
./Documentation/gpu/drm-kms:538: ./drivers/gpu/drm/drm_connector.c:2368: WARNING: Block quote ends without a blank line; unexpected unindent. [docutils]
./Documentation/gpu/drm-kms:538: ./drivers/gpu/drm/drm_connector.c:2381: ERROR: Unexpected indentation. [docutils]
Jani Nikula [Fri, 9 Aug 2024 14:27:06 +0000 (17:27 +0300)]
drm/i915/bios: convert to struct intel_display
Going forward, struct intel_display shall replace struct
drm_i915_private as the main display device data pointer type. Convert
intel_bios.[ch] to struct intel_display.
Do one drive-by conversion of unnecessary hex usage to decimal.
Jani Nikula [Fri, 9 Aug 2024 14:27:05 +0000 (17:27 +0300)]
drm/i915/opregion: convert to struct intel_display
Going forward, struct intel_display shall replace struct
drm_i915_private as the main display device data pointer type. Convert
intel_opregion.[ch] to struct intel_display.
v2:
- Fix declarations for !CONFIG_ACPI (Imre, kernel test robot)
- Pass encoder/connector directly to intel_display() (Imre)
Jani Nikula [Fri, 9 Aug 2024 14:27:03 +0000 (17:27 +0300)]
drm/i915/acpi: convert to struct intel_display
Going forward, struct intel_display shall replace struct
drm_i915_private as the main display device data pointer type. Convert
intel_acpi.[ch] to struct intel_display.
Mohammed Anees [Sun, 11 Aug 2024 10:16:51 +0000 (06:16 -0400)]
drm: Add missing documentation for struct drm_plane_size_hint
This patch takes care of the following warnings during documentation
compiling:
./include/uapi/drm/drm_mode.h:869: warning: Function parameter or struct member 'width' not described in 'drm_plane_size_hint'
./include/uapi/drm/drm_mode.h:869: warning: Function parameter or struct member 'height' not described in 'drm_plane_size_hint'
Jocelyn Falempe [Wed, 7 Aug 2024 13:36:14 +0000 (15:36 +0200)]
drm/panic: Add panic description
Now that kmsg dump callback has the description parameter, use it in
the user panic screen.
This is the string passed to panic(), like "VFS: Unable to mount root
fs on xxx" or "Attempted to kill init! exitcode=0xxxx".
It gives a hint on why the panic occurred, without being too cryptic.
Jocelyn Falempe [Wed, 7 Aug 2024 13:36:12 +0000 (15:36 +0200)]
drm/panic: Move drm_panic_register prototype to drm_crtc_internal.h
drm_panic_[un]register() are only used by the core drm, and are not
intended to be called by other drm drivers, so move their prototypes
to drm_crtc_internal.h.
projects / linux.git / log