summaryrefslogtreecommitdiff
path: root/drivers/gpu/drm/i915
AgeCommit message (Collapse)Author
2024-10-31drm/i915/dp: Write the source OUI for eDP before detecting sink capabilitiesImre Deak
The eDP sink's capabilities, like DSC, may depend on the source OUI written to the sink, so ensure the OUI is written before reading out the capabilities. Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20241025160259.3088727-7-imre.deak@intel.com
2024-10-31drm/i915/dp: Reuse intel_dp_detect_dsc_caps() for eDPImre Deak
Reuse intel_dp_detect_dsc_caps() which already checks for the source's DSC cap and retrieves the DPCD version from the DPRX caps. Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20241025160259.3088727-6-imre.deak@intel.com
2024-10-31drm/i915/dp: Track source OUI validity explicitlyImre Deak
While updating the source OUI on the sink the driver should avoid writing the OUI if it's already up-to-date to prevent the sink from resetting itself in response to the update. On eDP - the only output type where the OUI was updated so far - the driver ensured this by comparing the current source OUI DPCD register values with the expected Intel OUI value, skipping the update in case of a match. On some non-eDP sinks - at least on Synaptics branch devices - this method doesn't work, since the source OUI DPCD registers read back as all 0, even after updating the registers. Handle the above kind of sinks by tracking when the OUI was updated and so should be valid, regardless of what the DPCD registers contain. eDP sinks reset the written source OUI value when the panel power is disabled, invalidate the OUI state accordingly. This is required by a follow-up patch updating the source OUI for non-eDP sink types as well. v2: Fix setting intel_dp::oui_valid=true, if the DPCD register contains already the expected value. Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20241025160259.3088727-5-imre.deak@intel.com
2024-10-31drm/i915/dp: Initialize the source OUI write timestamp alwaysImre Deak
If the source OUI DPCD register value matches the expected Intel OUI value, the write timestamp doesn't get updated leaving it at the 0 initial value if the OUI wasn't written before. This can lead to an incorrect wait duration in intel_dp_wait_source_oui(), since jiffies is not inited to 0 in general (on a 32 bit system INITIAL_JIFFIES is set to 5 minutes ahead of wrap-around). Fix this by intializing the write timestamp in the above case as well. Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20241025160259.3088727-4-imre.deak@intel.com
2024-10-31drm/i915/dp: Ensure panel power remains enabled during connector detectionImre Deak
The sink's capabilities, like the DSC caps, depend on the source OUI written to the sink's DPCD registers and so this OUI value should be valid for the whole duration of the detection. An eDP sink will reset this OUI value when the panel power is disabled, so prevent the disabling - happening by default after a 1 sec idle period - for the whole duration of detection. v2: Update the documentation for intel_pps_on(). (Jani) Cc: Jani Nikula <jani.nikula@linux.intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20241025160259.3088727-3-imre.deak@intel.com
2024-10-31drm/i915/dp: Flush modeset commits during connector detectionImre Deak
Make sure that a DP connector detection doesn't happen in parallel with an ongoing modeset on the connector. The reasons for this are: - Besides reading the capabilities, EDID etc. the detection may change the state of the sink (via the AUX bus), for instance by setting the LTTPR mode or the source OUI (the latter introduced by an upcoming patch). It's better to avoid such changes affecting an onging modeset in any way. - During a modeset's link training any access to DPCD registers, besides the registers used for link training should be avoided, at least in the LTTPR non-transparent and transparent link training modes. Such asynchronous accesses - besides connector detection - can also happen via the AUX device node for instance, for those a parallel modeset will have to be avoided in a similar way to the change in this patch. (A topic for a follow-up change.) - The source OUI written to an eDP sink is valid only while the panel power is enabled. A modeset on eDP will enable/disable the panel power synchronously; this should be prevented in the middle of the connector detection, to ensure a consistent sink state (which depends on the source OUI) for the whole duration of detection. The panel power could still get disabled during detection after an idle period (1 sec), this will be prevented by the next patch. v2: (Ville) - s/wait_for_crtc_hw_done/wait_for_connector_hw_done - Get drm_device using an intel_display instead of drm_i915_private ptr. Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by: Imre Deak <imre.deak@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20241025160259.3088727-2-imre.deak@intel.com
2024-10-30drm/i915/display: use x100 version for full version and releaseJani Nikula
Use x100, or ver * 100 + rel, versions for full IP version checks, similar to what xe driver does: - Replace IP_VER(14, 1) inline with 1401, etc. - Convert DISPLAY_VER_FULL() to DISPLAY_VERx100() - Convert IS_DISPLAY_VER_FULL() to IS_DISPLAY_VERx100() - Convert IS_DISPLAY_VER_STEP() to IS_DISPLAY_VERx100_STEP() This makes ver.rel versions easier to use, follows the xe driver pattern, and drops the dependency on the IP_VER() macro. v2: Rebase, drop IP_VER() from xe compat headers v3: Rebase Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Acked-by: Matt Roper <matthew.d.roper@intel.com> Reviewed-by: Suraj Kandpal <suraj.kandpal@intel.com> Signed-off-by: Jani Nikula <jani.nikula@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20241029155536.753413-1-jani.nikula@intel.com
2024-10-30drm/i915/de: remove unnecessary generic wrappersJani Nikula
With many of the intel_de_* callers switched over to struct intel_display, we can remove some of the unnecessary generic wrappers. Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by: Jani Nikula <jani.nikula@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/82da66027a122b336278daa2c9a9eb39843082ba.1730146000.git.jani.nikula@intel.com
2024-10-30drm/i915/dsi: convert to struct intel_displayJani Nikula
struct intel_display will replace struct drm_i915_private as the main device pointer for display code. Switch ICL DSI code over to it. Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by: Jani Nikula <jani.nikula@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/f62a3616ef15e02cf19c5d041656fc6e09b37f6a.1730146000.git.jani.nikula@intel.com
2024-10-30drm/i915/ips: convert to struct intel_displayJani Nikula
struct intel_display will replace struct drm_i915_private as the main device pointer for display code. Switch HSW IPS code over to it. Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by: Jani Nikula <jani.nikula@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/66060d0c3fbb20e5d2c98a92133f091de6b25230.1730146000.git.jani.nikula@intel.com
2024-10-30drm/i915/power: convert assert_chv_phy_status() to struct intel_displayJani Nikula
struct intel_display will replace struct drm_i915_private as the main device pointer for display code. Switch assert_chv_phy_status() and its callers to it. Main motivation to do just one function is to stop passing i915 to intel_de_wait(), so its generic wrapper can be removed. Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by: Jani Nikula <jani.nikula@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/de6b01e1f21934ff520aa3b49ab5f97cbbf028f2.1730146000.git.jani.nikula@intel.com
2024-10-30drm/i915/display: convert vlv_wait_port_ready() to struct intel_displayJani Nikula
struct intel_display will replace struct drm_i915_private as the main device pointer for display code. Switch vlv_wait_port_ready() over to it. The main motivation to do just one function is to stop passing i915 to intel_de_wait(), so its generic wrapper can be removed. Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by: Jani Nikula <jani.nikula@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/9505ea49dfc8c7a52cacd2749875a680b01e5bbd.1730146000.git.jani.nikula@intel.com
2024-10-30drm/i915/crt: convert to struct intel_displayJani Nikula
struct intel_display will replace struct drm_i915_private as the main device pointer for display code. Switch CRT code over to it. v2: Rebase Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by: Jani Nikula <jani.nikula@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20241029105257.391572-1-jani.nikula@intel.com
2024-10-30drm/i915/dp/hdcp: convert to struct intel_displayJani Nikula
struct intel_display will replace struct drm_i915_private as the main device pointer for display code. Switch DP HDCP code over to it. v2: Rebase Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by: Jani Nikula <jani.nikula@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20241029090422.198749-1-jani.nikula@intel.com
2024-10-30drm/i915/hdcp: further conversion to struct intel_displayJani Nikula
There are some unconverted stragglers left in the HDCP API still using struct drm_i915_private. Convert to struct intel_display. Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by: Jani Nikula <jani.nikula@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/9680cc9e5ed7798a736fa73ad9ea0eb9c88e64bb.1730146000.git.jani.nikula@intel.com
2024-10-30drm/i915/dpio: convert to struct intel_displayJani Nikula
struct intel_display will replace struct drm_i915_private as the main device pointer for display code. Switch DPIO PHY code over to it. Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by: Jani Nikula <jani.nikula@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/1138083101f3c9058284592009b25f41065fbe30.1730146000.git.jani.nikula@intel.com
2024-10-30drm/i915/cx0: convert to struct intel_displayJani Nikula
struct intel_display will replace struct drm_i915_private as the main device pointer for display code. Switch Cx0 PHY code over to it. v2: Rebase, split out the include cleanups (Rodrigo) v3: Rebase Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by: Jani Nikula <jani.nikula@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20241029160822.800097-1-jani.nikula@intel.com
2024-10-30drm/i915/cx0: remove unnecessary includesJani Nikula
There's nothing in the header that requires the bit or bitfield headers. Remove. Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by: Jani Nikula <jani.nikula@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/3e12f1d5ab17e501e4700044072fbb6dd9b2f459.1730146000.git.jani.nikula@intel.com
2024-10-30drm/i915/gmbus: convert to struct intel_displayJani Nikula
struct intel_display will replace struct drm_i915_private as the main device pointer for display code. Switch gmbus code over to it. Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by: Jani Nikula <jani.nikula@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/d784e4799ab5095baa5c8fd840920066878c6273.1730146000.git.jani.nikula@intel.com
2024-10-30drm/i915/vblank: use display->platform.<platform> instead of IS_<PLATFORM>()Jani Nikula
Switch to using the new display->platform.<platform> members for platform identification in display code. v2: Split out an unrelated hunk to a separate patch (Rodrigo) Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/ed7295a7246bf00d8ae39f78c78dcc842c6939d9.1730144869.git.jani.nikula@intel.com Signed-off-by: Jani Nikula <jani.nikula@intel.com>
2024-10-30drm/i915/vblank: drop unnecessary i915 local variableJani Nikula
Use struct intel_display where possible. Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/c925fd0a56ba0a9183b62b7cc0ead0e37264f024.1730144869.git.jani.nikula@intel.com Signed-off-by: Jani Nikula <jani.nikula@intel.com>
2024-10-30drm/i915/vga: use display->platform.<platform> instead of IS_<PLATFORM>()Jani Nikula
Switch to using the new display->platform.<platform> members for platform identification in display code. Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/ee09546f217723a5f869749562eea7d6e97472f7.1730144869.git.jani.nikula@intel.com Signed-off-by: Jani Nikula <jani.nikula@intel.com>
2024-10-30drm/i915/tv: use display->platform.<platform> instead of IS_<PLATFORM>()Jani Nikula
Switch to using the new display->platform.<platform> members for platform identification in display code. Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/24306709d58bff03c819f44dfada95c1c998ad11.1730144869.git.jani.nikula@intel.com Signed-off-by: Jani Nikula <jani.nikula@intel.com>
2024-10-30drm/i915/pps: use display->platform.<platform> instead of IS_<PLATFORM>()Jani Nikula
Switch to using the new display->platform.<platform> members for platform identification in display code. Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/05de39ea5ef3898c115cc4f3725f58116d285339.1730144869.git.jani.nikula@intel.com Signed-off-by: Jani Nikula <jani.nikula@intel.com>
2024-10-30drm/i915/bios: use display->platform.<platform> instead of IS_<PLATFORM>()Jani Nikula
Switch to using the new display->platform.<platform> members for platform identification in display code. Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20dd28a25fbeedb36c576dfbbd11ec97376b903d.1730144869.git.jani.nikula@intel.com Signed-off-by: Jani Nikula <jani.nikula@intel.com>
2024-10-30drm/i915/display: add subplatform group for HSW/BDW ULTJani Nikula
Add support for defining aliases for subplatform groups, such as HSW/BDW ULT that covers both ULT and ULX. ULT is a special case, because we slightly abuse the ULT subplatform both as a subplatform and group, but with the way this is defined, it should be fairly clear. This follows i915 core and IS_HASWELL_ULT()/IS_BROADWELL_ULT() conventions, i.e. "is ULT" also matches ULX platforms. Note: Pedantically, this should have been done earlier, but it's only feasible now that we no longer have a subplatform enum and can actually initialize multiple subplatforms. v2: Use the subplatform group idea Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/d148b6210a561b874642ae3e0ad10073d0615de7.1730144869.git.jani.nikula@intel.com Signed-off-by: Jani Nikula <jani.nikula@intel.com>
2024-10-30drm/i915/display: add platform group for g4xJani Nikula
Add support for defining aliases for platform groups, such as g4x that covers both g45 and gm45. Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/947bdbc03913838383d75b3e07cf340100cbb5bb.1730144869.git.jani.nikula@intel.com Signed-off-by: Jani Nikula <jani.nikula@intel.com>
2024-10-30drm/i915/display: remove the display platform enum as unnecessaryJani Nikula
The display platform enums are not really needed for anything. Remove. Without the enum, PLATFORM_UNINITIALIZED is also no longer needed for keeping the first enum 0. Also need to switch from sp->subplatform to sp->pciidlist as the check for array end. Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/d20966f1d7a69a1e66768110b427be2fc611bcd2.1730144869.git.jani.nikula@intel.com Signed-off-by: Jani Nikula <jani.nikula@intel.com>
2024-10-30drm/i915/display: add platform member to struct intel_displayJani Nikula
Facilitate using display->platform.haswell and display->platform.haswell_ult etc. for identifying platforms and subplatforms. Merge the platform and subplatform bitmaps together, and check that there's no overlap. v4: - Lower case, s/is/platform/ v3: - Fix sanity check on display->is after merging subplatform members v2: - Use bitmap ops - Add some sanity checks with warnings Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/2ae79637390372903a9808b5adc4d2dcf2c5959b.1730144869.git.jani.nikula@intel.com Signed-off-by: Jani Nikula <jani.nikula@intel.com>
2024-10-30drm/i915/display: add display platforms structure with platform membersJani Nikula
Add a structure with a bitfield member for each platform and subplatform, and initialize them in platform and subplatform descs. The structure also contains a bitmap in a union for easier manipulation of the bits. This, in turn, requires a bit of trickery with INTEL_DISPLAY_PLATFORMS() to count the number of bits required for DECLARE_BITMAP(). Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/bf1d828cd333d34862ad3198e282c9d294c6e1ad.1730144869.git.jani.nikula@intel.com Signed-off-by: Jani Nikula <jani.nikula@intel.com>
2024-10-30drm/i915/display: convert display platforms to lower caseJani Nikula
This will be helpful for follow-up, where the names here become struct member names. This does impact debug logs as well, making everything lower case. v2: Rebase to adapt to PTL Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/8eab1be56093f33a7573e3caa78a4933bbf1ee76.1730144869.git.jani.nikula@intel.com Signed-off-by: Jani Nikula <jani.nikula@intel.com>
2024-10-30drm/i915/display: join the platform and subplatform enumsJani Nikula
We'll want to use the subplatforms similar to platforms. Join the subplatforms next to their corresponding platforms. Update the comment while at it. v2: Put the subplatforms next to the platforms Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/de274ffcd06a249a3983905b285c05d1c89953a8.1730144869.git.jani.nikula@intel.com Signed-off-by: Jani Nikula <jani.nikula@intel.com>
2024-10-30drm/i915/display: use a macro to define platform enumerationsJani Nikula
We'll be needing a macro based list of platforms for more things in the future. Start by defining the platform enumerations with it. v3: Rebase for PTL Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/fca14c018c54ef4099012d2a764257d651d672d9.1730144869.git.jani.nikula@intel.com Signed-off-by: Jani Nikula <jani.nikula@intel.com>
2024-10-30drm/i915/display: use a macro to initialize subplatformsJani Nikula
Make it easier to change the underlying structures by using a macro similar to PLATFORM() for initialization. The subplatform names in debug logs change slightly as they now reflect the enum rather than manually entered names. For example, RAPTORLAKE_S rather than RPL-S. Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/c4c9d7ea779475513db68e843c970a4dd8f8ac2c.1730144869.git.jani.nikula@intel.com Signed-off-by: Jani Nikula <jani.nikula@intel.com>
2024-10-30drm/i915/display: reindent subplatform initializationJani Nikula
Make the subplatform initialization less cramped, and follow the coding style more closely. Initialize .pciidlist using designated initializers. Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/0b7c3c44e4111c07e38a4bc842bbbced6f97c827.1730144869.git.jani.nikula@intel.com Signed-off-by: Jani Nikula <jani.nikula@intel.com>
2024-10-30fs: port files to file_refChristian Brauner
Port files to rely on file_ref reference to improve scaling and gain overflow protection. - We continue to WARN during get_file() in case a file that is already marked dead is revived as get_file() is only valid if the caller already holds a reference to the file. This hasn't changed just the check changes. - The semantics for epoll and ttm's dmabuf usage have changed. Both epoll and ttm synchronize with __fput() to prevent the underlying file from beeing freed. (1) epoll Explaining epoll is straightforward using a simple diagram. Essentially, the mutex of the epoll instance needs to be taken in both __fput() and around epi_fget() preventing the file from being freed while it is polled or preventing the file from being resurrected. CPU1 CPU2 fput(file) -> __fput(file) -> eventpoll_release(file) -> eventpoll_release_file(file) mutex_lock(&ep->mtx) epi_item_poll() -> epi_fget() -> file_ref_get(file) mutex_unlock(&ep->mtx) mutex_lock(&ep->mtx); __ep_remove() mutex_unlock(&ep->mtx); -> kmem_cache_free(file) (2) ttm dmabuf This explanation is a bit more involved. A regular dmabuf file stashed the dmabuf in file->private_data and the file in dmabuf->file: file->private_data = dmabuf; dmabuf->file = file; The generic release method of a dmabuf file handles file specific things: f_op->release::dma_buf_file_release() while the generic dentry release method of a dmabuf handles dmabuf freeing including driver specific things: dentry->d_release::dma_buf_release() During ttm dmabuf initialization in ttm_object_device_init() the ttm driver copies the provided struct dma_buf_ops into a private location: struct ttm_object_device { spinlock_t object_lock; struct dma_buf_ops ops; void (*dmabuf_release)(struct dma_buf *dma_buf); struct idr idr; }; ttm_object_device_init(const struct dma_buf_ops *ops) { // copy original dma_buf_ops in private location tdev->ops = *ops; // stash the release method of the original struct dma_buf_ops tdev->dmabuf_release = tdev->ops.release; // override the release method in the copy of the struct dma_buf_ops // with ttm's own dmabuf release method tdev->ops.release = ttm_prime_dmabuf_release; } When a new dmabuf is created the struct dma_buf_ops with the overriden release method set to ttm_prime_dmabuf_release is passed in exp_info.ops: DEFINE_DMA_BUF_EXPORT_INFO(exp_info); exp_info.ops = &tdev->ops; exp_info.size = prime->size; exp_info.flags = flags; exp_info.priv = prime; The call to dma_buf_export() then sets mutex_lock_interruptible(&prime->mutex); dma_buf = dma_buf_export(&exp_info) { dmabuf->ops = exp_info->ops; } mutex_unlock(&prime->mutex); which creates a new dmabuf file and then install a file descriptor to it in the callers file descriptor table: ret = dma_buf_fd(dma_buf, flags); When that dmabuf file is closed we now get: fput(file) -> __fput(file) -> f_op->release::dma_buf_file_release() -> dput() -> d_op->d_release::dma_buf_release() -> dmabuf->ops->release::ttm_prime_dmabuf_release() mutex_lock(&prime->mutex); if (prime->dma_buf == dma_buf) prime->dma_buf = NULL; mutex_unlock(&prime->mutex); Where we can see that prime->dma_buf is set to NULL. So when we have the following diagram: CPU1 CPU2 fput(file) -> __fput(file) -> f_op->release::dma_buf_file_release() -> dput() -> d_op->d_release::dma_buf_release() -> dmabuf->ops->release::ttm_prime_dmabuf_release() ttm_prime_handle_to_fd() mutex_lock_interruptible(&prime->mutex) dma_buf = prime->dma_buf dma_buf && get_dma_buf_unless_doomed(dma_buf) -> file_ref_get(dma_buf->file) mutex_unlock(&prime->mutex); mutex_lock(&prime->mutex); if (prime->dma_buf == dma_buf) prime->dma_buf = NULL; mutex_unlock(&prime->mutex); -> kmem_cache_free(file) The logic of the mechanism is the same as for epoll: sync with __fput() preventing the file from being freed. Here the synchronization happens through the ttm instance's prime->mutex. Basically, the lifetime of the dma_buf and the file are tighly coupled. Both (1) and (2) used to call atomic_inc_not_zero() to check whether the file has already been marked dead and then refuse to revive it. This is only safe because both (1) and (2) sync with __fput() and thus prevent kmem_cache_free() on the file being called and thus prevent the file from being immediately recycled due to SLAB_TYPESAFE_BY_RCU. Both (1) and (2) have been ported from atomic_inc_not_zero() to file_ref_get(). That means a file that is already in the process of being marked as FILE_REF_DEAD: file_ref_put() cnt = atomic_long_dec_return() -> __file_ref_put(cnt) if (cnt == FIlE_REF_NOREF) atomic_long_try_cmpxchg_release(cnt, FILE_REF_DEAD) can be revived again: CPU1 CPU2 file_ref_put() cnt = atomic_long_dec_return() -> __file_ref_put(cnt) if (cnt == FIlE_REF_NOREF) file_ref_get() // Brings reference back to FILE_REF_ONEREF atomic_long_add_negative() atomic_long_try_cmpxchg_release(cnt, FILE_REF_DEAD) This is fine and inherent to the file_ref_get()/file_ref_put() semantics. For both (1) and (2) this is safe because __fput() is prevented from making progress if file_ref_get() fails due to the aforementioned synchronization mechanisms. Two cases need to be considered that affect both (1) epoll and (2) ttm dmabuf: (i) fput()'s file_ref_put() and marks the file as FILE_REF_NOREF but before that fput() can mark the file as FILE_REF_DEAD someone manages to sneak in a file_ref_get() and brings the refcount back from FILE_REF_NOREF to FILE_REF_ONEREF. In that case the original fput() doesn't call __fput(). For epoll the poll will finish and for ttm dmabuf the file can be used again. For ttm dambuf this is actually an advantage because it avoids immediately allocating a new dmabuf object. CPU1 CPU2 file_ref_put() cnt = atomic_long_dec_return() -> __file_ref_put(cnt) if (cnt == FIlE_REF_NOREF) file_ref_get() // Brings reference back to FILE_REF_ONEREF atomic_long_add_negative() atomic_long_try_cmpxchg_release(cnt, FILE_REF_DEAD) (ii) fput()'s file_ref_put() marks the file FILE_REF_NOREF and also suceeds in actually marking it FILE_REF_DEAD and then calls into __fput() to free the file. When either (1) or (2) call file_ref_get() they fail as atomic_long_add_negative() will return true. At the same time, both (1) and (2) all file_ref_get() under mutexes that __fput() must also acquire preventing kmem_cache_free() from freeing the file. So while this might be treated as a change in semantics for (1) and (2) it really isn't. It if should end up causing issues this can be fixed by adding a helper that does something like: long cnt = atomic_long_read(&ref->refcnt); do { if (cnt < 0) return false; } while (!atomic_long_try_cmpxchg(&ref->refcnt, &cnt, cnt + 1)); return true; which would block FILE_REF_NOREF to FILE_REF_ONEREF transitions. - Jann correctly pointed out that kmem_cache_zalloc() cannot be used anymore once files have been ported to file_ref_t. The kmem_cache_zalloc() call will memset() the whole struct file to zero when it is reallocated. This will also set file->f_ref to zero which mens that a concurrent file_ref_get() can return true: CPU1 CPU2 __get_file_rcu() rcu_dereference_raw() close() [frees file] alloc_empty_file() kmem_cache_zalloc() [reallocates same file] memset(..., 0, ...) file_ref_get() [increments 0->1, returns true] init_file() file_ref_init(..., 1) [sets to 0] rcu_dereference_raw() fput() file_ref_put() [decrements 0->FILE_REF_NOREF, frees file] [UAF] causing a concurrent __get_file_rcu() call to acquire a reference to the file that is about to be reallocated and immediately freeing it on realizing that it has been recycled. This causes a UAF for the task that reallocated/recycled the file. This is prevented by switching from kmem_cache_zalloc() to kmem_cache_alloc() and initializing the fields manually. With file->f_ref initialized last. Note that a memset() also isn't guaranteed to atomically update an unsigned long so it's theoretically possible to see torn and therefore bogus counter values. Link: https://lore.kernel.org/r/20241007-brauner-file-rcuref-v2-3-387e24dc9163@kernel.org Signed-off-by: Christian Brauner <brauner@kernel.org>
2024-10-29drm/i915/display/dp: Reduce log level for SOURCE OUI write failuresClint Taylor
Some devices NAK DPCD writes to the SOURCE OUI (0x300) DPCD registers. Reduce the log level priority to prevent dmesg noise for these devices. Signed-off-by: Clint Taylor <clinton.a.taylor@intel.com> Reviewed-by: Sai Teja Pottumuttu <sai.teja.pottumuttu@intel.com> Signed-off-by: Matt Roper <matthew.d.roper@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20241004210816.3976058-1-clinton.a.taylor@intel.com
2024-10-29drm/i915/color: Make color .get_config() mandatoryVille Syrjälä
Every platforms implements the color .get_config() hook. Just make it mandatory. Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20241024165356.17756-5-ville.syrjala@linux.intel.com Reviewed-by: Jani Nikula <jani.nikula@intel.com>
2024-10-29drm/i915/color: Convert color management code to intel_displayVille Syrjälä
struct intel_display will replace struct drm_i915_private as the main thing for display code. Convert the color management code to use it (as much as possible at this stage). Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20241024165356.17756-4-ville.syrjala@linux.intel.com Reviewed-by: Jani Nikula <jani.nikula@intel.com>
2024-10-29drm/i915: Handle intel_plane and intel_plane_state in to_intel_display()Ville Syrjälä
Allow one to pass intel_plane/intel_plane_state to to_intel_display(). Works exactly like their crtc counterparts. Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20241024165356.17756-3-ville.syrjala@linux.intel.com Reviewed-by: Jani Nikula <jani.nikula@intel.com>
2024-10-29drm/i915/color: Pimp debugsVille Syrjälä
Include the CRTC id+name information in the color management debug prints to help identify who is at fault. And also specify which LUT check_lut_size() is unhappy about. Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20241024165356.17756-2-ville.syrjala@linux.intel.com Reviewed-by: Jani Nikula <jani.nikula@intel.com>
2024-10-29drm/i915/display/xe3: disable x-tiled framebuffersHeikkila, Juha-pekka
Xe3 has no more support for x-tile on display. v2: Include up to display 29 for X-tiled support. (Gustavo) Signed-off-by: Heikkila, Juha-pekka <juha-pekka.heikkila@intel.com> Signed-off-by: Matt Atwood <matthew.s.atwood@intel.com> Signed-off-by: Clint Taylor <Clinton.A.Taylor@intel.com> Reviewed-by: Gustavo Sousa <gustavo.sousa@intel.com> Signed-off-by: Matt Roper <matthew.d.roper@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20241028193015.3241858-9-clinton.a.taylor@intel.com
2024-10-29drm/i915/xe3: Underrun recovery does not exist post Xe2Ravi Kumar Vodapalli
From platforms xe3 Underrun recovery does not exist v2: improve DISPLAY_VER checking BSpec: 68849 Signed-off-by: Ravi Kumar Vodapalli <ravi.kumar.vodapalli@intel.com> Signed-off-by: Matt Atwood <matthew.s.atwood@intel.com> Signed-off-by: Clint Taylor <Clinton.A.Taylor@intel.com> Reviewed-by: Sai Teja Pottumuttu <sai.teja.pottumuttu@intel.com> Signed-off-by: Matt Roper <matthew.d.roper@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20241028193015.3241858-8-clinton.a.taylor@intel.com
2024-10-29drm/i915/xe3lpd: Move async flip bit to PLANE_SURF registerDnyaneshwar Bhadane
The async flip moved from PLANE_CTL to PLANE_SURF for Xe3_LPD. Bspec: 69853,69878 Signed-off-by: Dnyaneshwar Bhadane <dnyaneshwar.bhadane@intel.com> Signed-off-by: Matt Atwood <matthew.s.atwood@intel.com> Signed-off-by: Clint Taylor <Clinton.A.Taylor@intel.com> Reviewed-by: Shekhar Chauhan <shekhar.chauhan@intel.com> Signed-off-by: Matt Roper <matthew.d.roper@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20241028193015.3241858-7-clinton.a.taylor@intel.com
2024-10-29drm/i915/cx0: Remove bus reset after every c10 transactionClint Taylor
C10 phy timeouts occur on xe3lpd if the c10 bus is reset every transaction. Although not required by BSPEC bus resets were added for prior platforms as a workaround. Starting with xe3_lpd this bus reset is not necessary. Signed-off-by: Clint Taylor <clinton.a.taylor@intel.com> Reviewed-by: Gustavo Sousa <gustavo.sousa@intel.com> Signed-off-by: Matt Roper <matthew.d.roper@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20241028193015.3241858-6-clinton.a.taylor@intel.com
2024-10-29drm/i915/cx0: Extend C10 check to PTLDnyaneshwar Bhadane
When deciding the type of the phy, add PTL support to make sure the correct path is taken for selection of C10 PHY. Only port A is connected C10 PHY for Pantherlake. Bspec: 72571 Signed-off-by: Dnyaneshwar Bhadane <dnyaneshwar.bhadane@intel.com> Signed-off-by: Matt Atwood <matthew.s.atwood@intel.com> Signed-off-by: Clint Taylor <Clinton.A.Taylor@intel.com> Reviewed-by: Gustavo Sousa <gustavo.sousa@intel.com> Signed-off-by: Matt Roper <matthew.d.roper@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20241028193015.3241858-5-clinton.a.taylor@intel.com
2024-10-29drm/i915/ptl: Define IS_PANTHERLAKE macroDnyaneshwar Bhadane
Common display code requires IS_PANTHERLAKE macro. Define the macro and set 0 as PTL is no longer support for i915. Signed-off-by: Dnyaneshwar Bhadane <dnyaneshwar.bhadane@intel.com> Signed-off-by: Matt Atwood <matthew.s.atwood@intel.com> Signed-off-by: Clint Taylor <Clinton.A.Taylor@intel.com> Reviewed-by: Matt Roper <matthew.d.roper@intel.com> Signed-off-by: Matt Roper <matthew.d.roper@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20241028193015.3241858-4-clinton.a.taylor@intel.com
2024-10-29drm/i915/xe3lpd: Add check to see if edp over type c is allowedSuraj Kandpal
Read PICA register to see if edp over type C is possible and then add the appropriate tables for it. --v2 -remove bool from intel_encoder have it in runtime_info [Jani] -initialize the bool in runtime_info init [Jani] -dont abbreviate the bool [Jani] --v3 -Remove useless display version check [Jani] -change the warn on condition [Jani] -no need for a different function for edp type c check [Jani] -dont add register in i915_reg [Jani] Bspec: 68846 Signed-off-by: Suraj Kandpal <suraj.kandpal@intel.com> Signed-off-by: Matt Atwood <matthew.s.atwood@intel.com> Signed-off-by: Clint Taylor <clinton.a.taylor@intel.com> Reviewed-by: Arun R Murthy <arun.r.murthy@intel.com> Acked-by: Jani Nikula <jani.nikula@intel.com> Signed-off-by: Matt Roper <matthew.d.roper@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20241028193015.3241858-3-clinton.a.taylor@intel.com
2024-10-29drm/i915/xe3lpd: Update pmdemand programmingMatt Roper
There are some minor changes to pmdemand handling on Xe3: - Active scalers are no longer tracked. We can simply skip the readout and programming of this field. - Active dbuf slices are no longer tracked. We should skip the readout and programming of this field and also make sure that it stays 0 in our software bookkeeping so that we won't erroneously return true from intel_pmdemand_needs_update() due to mismatches. - Even though there aren't enough pipes to utilize them, the size of the 'active pipes' field has expanded to four bits, taking over the register bits previously used for dbuf slices. Since the lower bits of the mask have moved, we need to update our reads/writes to handle this properly. v2: active pipes is no longer always max 3, add in the ability to go to 4 for PTL. v3: use intel_display for display_ver check, use INTEL_NUM_PIPES v4: add a conditional for number of pipes macro vs using 3. v5: reverse conditional order of v4. v6: undo v5 and fix num_pipes assignment v7: pass display struct instead of i915, checkpatch fix v8: Alignment issue Bspec: 68883, 69125 Signed-off-by: Matt Roper <matthew.d.roper@intel.com> Signed-off-by: Matt Atwood <matthew.s.atwood@intel.com> Signed-off-by: Clint Taylor <Clinton.A.Taylor@intel.com> Reviewed-by: Vinod Govindapillai <vinod.govindapillai@intel.com> Reviewed-by: Gustavo Sousa <gustavo.sousa@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20241028193015.3241858-2-clinton.a.taylor@intel.com
2024-10-29drm/intel/pciids: rename i915_pciids.h to just pciids.hJani Nikula
In preparation of sharing the PCI ID macros between i915 and xe, rename i915_pciids.h to pciids.h. Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com> Cc: Lucas De Marchi <lucas.demarchi@intel.com> Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Cc: Thomas Hellström <thomas.hellstrom@linux.intel.com> Cc: Tvrtko Ursulin <tursulin@ursulin.net> Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com> Acked-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/835143845faa5310e4bb58405a8a0848392bbf06.1729590029.git.jani.nikula@intel.com Signed-off-by: Jani Nikula <jani.nikula@intel.com>
generated by cgit (git 2.34.1) at 2025-09-14 21:43:19 +0000