From eb0851e14432f3b87c77b704c835ac376deda03a Mon Sep 17 00:00:00 2001 From: I Hsin Cheng Date: Tue, 6 May 2025 02:43:38 +0800 Subject: [PATCH 01/24] drm/meson: Use 1000ULL when operating with mode->clock Coverity scan reported the usage of "mode->clock * 1000" may lead to integer overflow. Use "1000ULL" instead of "1000" when utilizing it to avoid potential integer overflow issue. Link: https://scan5.scan.coverity.com/#/project-view/10074/10063?selectedIssue=1646759 Signed-off-by: I Hsin Cheng Reviewed-by: Martin Blumenstingl Fixes: 1017560164b6 ("drm/meson: use unsigned long long / Hz for frequency types") Link: https://lore.kernel.org/r/20250505184338.678540-1-richard120310@gmail.com Signed-off-by: Neil Armstrong --- drivers/gpu/drm/meson/meson_encoder_hdmi.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/meson/meson_encoder_hdmi.c b/drivers/gpu/drm/meson/meson_encoder_hdmi.c index 7752d8ac85f0..c08fa93e50a3 100644 --- a/drivers/gpu/drm/meson/meson_encoder_hdmi.c +++ b/drivers/gpu/drm/meson/meson_encoder_hdmi.c @@ -75,7 +75,7 @@ static void meson_encoder_hdmi_set_vclk(struct meson_encoder_hdmi *encoder_hdmi, unsigned long long venc_freq; unsigned long long hdmi_freq; - vclk_freq = mode->clock * 1000; + vclk_freq = mode->clock * 1000ULL; /* For 420, pixel clock is half unlike venc clock */ if (encoder_hdmi->output_bus_fmt == MEDIA_BUS_FMT_UYYVYY8_0_5X24) @@ -123,7 +123,7 @@ static enum drm_mode_status meson_encoder_hdmi_mode_valid(struct drm_bridge *bri struct meson_encoder_hdmi *encoder_hdmi = bridge_to_meson_encoder_hdmi(bridge); struct meson_drm *priv = encoder_hdmi->priv; bool is_hdmi2_sink = display_info->hdmi.scdc.supported; - unsigned long long clock = mode->clock * 1000; + unsigned long long clock = mode->clock * 1000ULL; unsigned long long phy_freq; unsigned long long vclk_freq; unsigned long long venc_freq; From 9c1798259b9420f38f1fa1b83e3d864c3eb1a83e Mon Sep 17 00:00:00 2001 From: Fabio Estevam Date: Thu, 17 Apr 2025 07:34:58 -0300 Subject: [PATCH 02/24] drm/tiny: panel-mipi-dbi: Use drm_client_setup_with_fourcc() Since commit 559358282e5b ("drm/fb-helper: Don't use the preferred depth for the BPP default"), RGB565 displays such as the CFAF240320X no longer render correctly: colors are distorted and the content is shown twice horizontally. This regression is due to the fbdev emulation layer defaulting to 32 bits per pixel, whereas the display expects 16 bpp (RGB565). As a result, the framebuffer data is incorrectly interpreted by the panel. Fix the issue by calling drm_client_setup_with_fourcc() with a format explicitly selected based on the display's bits-per-pixel value. For 16 bpp, use DRM_FORMAT_RGB565; for other values, fall back to the previous behavior. This ensures that the allocated framebuffer format matches the hardware expectations, avoiding color and layout corruption. Tested on a CFAF240320X display with an RGB565 configuration, confirming correct colors and layout after applying this patch. Cc: stable@vger.kernel.org Fixes: 559358282e5b ("drm/fb-helper: Don't use the preferred depth for the BPP default") Signed-off-by: Fabio Estevam Reviewed-by: Thomas Zimmermann Reviewed-by: Javier Martinez Canillas Signed-off-by: Thomas Zimmermann Link: https://lore.kernel.org/r/20250417103458.2496790-1-festevam@gmail.com --- drivers/gpu/drm/tiny/panel-mipi-dbi.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/tiny/panel-mipi-dbi.c b/drivers/gpu/drm/tiny/panel-mipi-dbi.c index 0460ecaef4bd..23914a9f7fd3 100644 --- a/drivers/gpu/drm/tiny/panel-mipi-dbi.c +++ b/drivers/gpu/drm/tiny/panel-mipi-dbi.c @@ -390,7 +390,10 @@ static int panel_mipi_dbi_spi_probe(struct spi_device *spi) spi_set_drvdata(spi, drm); - drm_client_setup(drm, NULL); + if (bpp == 16) + drm_client_setup_with_fourcc(drm, DRM_FORMAT_RGB565); + else + drm_client_setup_with_fourcc(drm, DRM_FORMAT_RGB888); return 0; } From 2f0c6fceae0df59a9c0edb5d41529b90ee9f2fcb Mon Sep 17 00:00:00 2001 From: Markus Burri Date: Thu, 8 May 2025 15:06:08 +0200 Subject: [PATCH 03/24] accel/ivpu: Use effective buffer size for zero terminator Use the effective written size instead of original size as index for zero termination. If the input from user-space is to larger and the input is truncated, the original size is out-of-bound. Since there is an upfront size check here, the change is for consistency. Signed-off-by: Markus Burri Reviewed-by: Jacek Lawrynowicz Signed-off-by: Jacek Lawrynowicz Link: https://lore.kernel.org/r/20250508130612.82270-3-markus.burri@mt.com --- drivers/accel/ivpu/ivpu_debugfs.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/accel/ivpu/ivpu_debugfs.c b/drivers/accel/ivpu/ivpu_debugfs.c index f0dad0c9ce33..cd24ccd20ba6 100644 --- a/drivers/accel/ivpu/ivpu_debugfs.c +++ b/drivers/accel/ivpu/ivpu_debugfs.c @@ -455,7 +455,7 @@ priority_bands_fops_write(struct file *file, const char __user *user_buf, size_t if (ret < 0) return ret; - buf[size] = '\0'; + buf[ret] = '\0'; ret = sscanf(buf, "%u %u %u %u", &band, &grace_period, &process_grace_period, &process_quantum); if (ret != 4) From a0fa7873f2f869087b1e7793f7fac3713a1e3afe Mon Sep 17 00:00:00 2001 From: Philip Yang Date: Wed, 7 May 2025 11:04:32 -0400 Subject: [PATCH 04/24] drm/amdgpu: csa unmap use uninterruptible lock MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit After process exit to unmap csa and free GPU vm, if signal is accepted and then waiting to take vm lock is interrupted and return, it causes memory leaking and below warning backtrace. Change to use uninterruptible wait lock fix the issue. WARNING: CPU: 69 PID: 167800 at amd/amdgpu/amdgpu_kms.c:1525 amdgpu_driver_postclose_kms+0x294/0x2a0 [amdgpu] Call Trace: drm_file_free.part.0+0x1da/0x230 [drm] drm_close_helper.isra.0+0x65/0x70 [drm] drm_release+0x6a/0x120 [drm] amdgpu_drm_release+0x51/0x60 [amdgpu] __fput+0x9f/0x280 ____fput+0xe/0x20 task_work_run+0x67/0xa0 do_exit+0x217/0x3c0 do_group_exit+0x3b/0xb0 get_signal+0x14a/0x8d0 arch_do_signal_or_restart+0xde/0x100 exit_to_user_mode_loop+0xc1/0x1a0 exit_to_user_mode_prepare+0xf4/0x100 syscall_exit_to_user_mode+0x17/0x40 do_syscall_64+0x69/0xc0 Signed-off-by: Philip Yang Reviewed-by: Christian König Signed-off-by: Alex Deucher (cherry picked from commit 7dbbfb3c171a6f63b01165958629c9c26abf38ab) Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/amdgpu/amdgpu_csa.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_csa.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_csa.c index cfdf558b48b6..02138aa55793 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_csa.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_csa.c @@ -109,7 +109,7 @@ int amdgpu_unmap_static_csa(struct amdgpu_device *adev, struct amdgpu_vm *vm, struct drm_exec exec; int r; - drm_exec_init(&exec, DRM_EXEC_INTERRUPTIBLE_WAIT, 0); + drm_exec_init(&exec, 0, 0); drm_exec_until_all_locked(&exec) { r = amdgpu_vm_lock_pd(vm, &exec, 0); if (likely(!r)) From 2d73b0845ab3963856e857b810600e5594bc29f4 Mon Sep 17 00:00:00 2001 From: Tim Huang Date: Thu, 8 May 2025 13:37:35 +0800 Subject: [PATCH 05/24] drm/amdgpu: fix incorrect MALL size for GFX1151 On GFX1151, the reported MALL cache size reflects only half of its actual size; this adjustment corrects the discrepancy. Signed-off-by: Tim Huang Acked-by: Alex Deucher Reviewed-by: Yifan Zhang Signed-off-by: Alex Deucher (cherry picked from commit 0a5c060b593ad152318f89e5564bfdfcff8a6ac0) Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c | 12 ++++++++++++ 1 file changed, 12 insertions(+) diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c index e74e26b6a4f2..fec9a007533a 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c @@ -752,6 +752,18 @@ static int gmc_v11_0_sw_init(struct amdgpu_ip_block *ip_block) adev->gmc.vram_type = vram_type; adev->gmc.vram_vendor = vram_vendor; + /* The mall_size is already calculated as mall_size_per_umc * num_umc. + * However, for gfx1151, which features a 2-to-1 UMC mapping, + * the result must be multiplied by 2 to determine the actual mall size. + */ + switch (amdgpu_ip_version(adev, GC_HWIP, 0)) { + case IP_VERSION(11, 5, 1): + adev->gmc.mall_size *= 2; + break; + default: + break; + } + switch (amdgpu_ip_version(adev, GC_HWIP, 0)) { case IP_VERSION(11, 0, 0): case IP_VERSION(11, 0, 1): From d433981385c62c72080e26f1c00a961d18b233be Mon Sep 17 00:00:00 2001 From: Wayne Lin Date: Fri, 25 Apr 2025 14:44:02 +0800 Subject: [PATCH 06/24] drm/amd/display: Correct the reply value when AUX write incomplete [Why] Now forcing aux->transfer to return 0 when incomplete AUX write is inappropriate. It should return bytes have been transferred. [How] aux->transfer is asked not to change original msg except reply field of drm_dp_aux_msg structure. Copy the msg->buffer when it's write request, and overwrite the first byte when sink reply 1 byte indicating partially written byte number. Then we can return the correct value without changing the original msg. Fixes: 3637e457eb00 ("drm/amd/display: Fix wrong handling for AUX_DEFER case") Cc: Mario Limonciello Cc: Alex Deucher Reviewed-by: Ray Wu Signed-off-by: Wayne Lin Signed-off-by: Ray Wu Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher (cherry picked from commit 7ac37f0dcd2e0b729fa7b5513908dc8ab802b540) Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 3 ++- .../drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c | 10 ++++++++-- 2 files changed, 10 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c index 64df8ca448b3..1525c408d452 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@ -12763,7 +12763,8 @@ int amdgpu_dm_process_dmub_aux_transfer_sync( /* The reply is stored in the top nibble of the command. */ payload->reply[0] = (adev->dm.dmub_notify->aux_reply.command >> 4) & 0xF; - if (!payload->write && p_notify->aux_reply.length) + /*write req may receive a byte indicating partially written number as well*/ + if (p_notify->aux_reply.length) memcpy(payload->data, p_notify->aux_reply.data, p_notify->aux_reply.length); diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c index 074b79fd5822..c93eef6f7a6b 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c @@ -62,6 +62,7 @@ static ssize_t dm_dp_aux_transfer(struct drm_dp_aux *aux, enum aux_return_code_type operation_result; struct amdgpu_device *adev; struct ddc_service *ddc; + uint8_t copy[16]; if (WARN_ON(msg->size > 16)) return -E2BIG; @@ -77,6 +78,11 @@ static ssize_t dm_dp_aux_transfer(struct drm_dp_aux *aux, (msg->request & DP_AUX_I2C_WRITE_STATUS_UPDATE) != 0; payload.defer_delay = 0; + if (payload.write) { + memcpy(copy, msg->buffer, msg->size); + payload.data = copy; + } + result = dc_link_aux_transfer_raw(TO_DM_AUX(aux)->ddc_service, &payload, &operation_result); @@ -100,9 +106,9 @@ static ssize_t dm_dp_aux_transfer(struct drm_dp_aux *aux, */ if (payload.write && result >= 0) { if (result) { - /*one byte indicating partially written bytes. Force 0 to retry*/ + /*one byte indicating partially written bytes*/ drm_info(adev_to_drm(adev), "amdgpu: AUX partially written\n"); - result = 0; + result = payload.data[0]; } else if (!payload.reply[0]) /*I2C_ACK|AUX_ACK*/ result = msg->size; From 190818d1b653fbf16138286345e67c8c1a83f29f Mon Sep 17 00:00:00 2001 From: Gabe Teeger Date: Fri, 25 Apr 2025 10:31:39 -0400 Subject: [PATCH 07/24] Revert: "drm/amd/display: Enable urgent latency adjustment on DCN35" This reverts commit 756c85e4d0dd ("drm/amd/display: Enable urgent latency adjustment on DCN35") Reason for revert: Negative power impact. Reviewed-by: Nicholas Kazlauskas Signed-off-by: Gabe Teeger Signed-off-by: Ray Wu Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher (cherry picked from commit 9334c491cd8f388232b9a187bf0ddb728482bd6f) --- drivers/gpu/drm/amd/display/dc/dml/dcn35/dcn35_fpu.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn35/dcn35_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn35/dcn35_fpu.c index d9159ca55412..92f0a099d089 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn35/dcn35_fpu.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn35/dcn35_fpu.c @@ -195,9 +195,9 @@ struct _vcs_dpi_soc_bounding_box_st dcn3_5_soc = { .dcn_downspread_percent = 0.5, .gpuvm_min_page_size_bytes = 4096, .hostvm_min_page_size_bytes = 4096, - .do_urgent_latency_adjustment = 1, + .do_urgent_latency_adjustment = 0, .urgent_latency_adjustment_fabric_clock_component_us = 0, - .urgent_latency_adjustment_fabric_clock_reference_mhz = 3000, + .urgent_latency_adjustment_fabric_clock_reference_mhz = 0, }; void dcn35_build_wm_range_table_fpu(struct clk_mgr *clk_mgr) From 874697e127931bf50a37ce9d96ee80f3a08a0c38 Mon Sep 17 00:00:00 2001 From: John Olender Date: Wed, 16 Apr 2025 02:54:26 -0400 Subject: [PATCH 08/24] drm/amd/display: Defer BW-optimization-blocked DRR adjustments [Why & How] Instead of dropping DRR updates, defer them. This fixes issues where monitor continues to see incorrect refresh rate after VRR was turned off by userspace. Fixes: 32953485c558 ("drm/amd/display: Do not update DRR while BW optimizations pending") Link: https://gitlab.freedesktop.org/drm/amd/-/issues/3546 Reviewed-by: Sun peng Li Signed-off-by: John Olender Signed-off-by: Aurabindo Pillai Signed-off-by: Ray Wu Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher (cherry picked from commit 53761b7ecd83e6fbb9f2206f8c980a6aa308c844) Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 2 ++ drivers/gpu/drm/amd/display/dc/core/dc.c | 10 +++++++--- 2 files changed, 9 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c index 1525c408d452..cc01b9c68b47 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@ -372,6 +372,8 @@ get_crtc_by_otg_inst(struct amdgpu_device *adev, static inline bool is_dc_timing_adjust_needed(struct dm_crtc_state *old_state, struct dm_crtc_state *new_state) { + if (new_state->stream->adjust.timing_adjust_pending) + return true; if (new_state->freesync_config.state == VRR_STATE_ACTIVE_FIXED) return true; else if (amdgpu_dm_crtc_vrr_active(old_state) != amdgpu_dm_crtc_vrr_active(new_state)) diff --git a/drivers/gpu/drm/amd/display/dc/core/dc.c b/drivers/gpu/drm/amd/display/dc/core/dc.c index 28d1353f403d..ba4ce8a63158 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc.c @@ -439,9 +439,12 @@ bool dc_stream_adjust_vmin_vmax(struct dc *dc, * Don't adjust DRR while there's bandwidth optimizations pending to * avoid conflicting with firmware updates. */ - if (dc->ctx->dce_version > DCE_VERSION_MAX) - if (dc->optimized_required || dc->wm_optimized_required) + if (dc->ctx->dce_version > DCE_VERSION_MAX) { + if (dc->optimized_required || dc->wm_optimized_required) { + stream->adjust.timing_adjust_pending = true; return false; + } + } dc_exit_ips_for_hw_access(dc); @@ -3168,7 +3171,8 @@ static void copy_stream_update_to_stream(struct dc *dc, if (update->crtc_timing_adjust) { if (stream->adjust.v_total_min != update->crtc_timing_adjust->v_total_min || - stream->adjust.v_total_max != update->crtc_timing_adjust->v_total_max) + stream->adjust.v_total_max != update->crtc_timing_adjust->v_total_max || + stream->adjust.timing_adjust_pending) update->crtc_timing_adjust->timing_adjust_pending = true; stream->adjust = *update->crtc_timing_adjust; update->crtc_timing_adjust->timing_adjust_pending = false; From 3c1a467372e0c356b1d3c59f6d199ed5a6612dd1 Mon Sep 17 00:00:00 2001 From: George Shen Date: Thu, 24 Apr 2025 10:02:59 -0400 Subject: [PATCH 09/24] drm/amd/display: fix link_set_dpms_off multi-display MST corner case [Why & How] When MST config is unplugged/replugged too quickly, it can potentially result in a scenario where previous DC state has not been reset before the HPD link detection sequence begins. In this case, driver will disable the streams/link prior to re-enabling the link for link training. There is a bug in the current logic that does not account for the fact that current_state can be released and cleared prior to swapping to a new state (resulting in the pipe_ctx stream pointers to be cleared) in between disabling streams. To resolve this, cache the original streams prior to committing any stream updates. Reviewed-by: Wenjing Liu Signed-off-by: George Shen Signed-off-by: Ray Wu Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher (cherry picked from commit 1561782686ccc36af844d55d31b44c938dd412dc) --- drivers/gpu/drm/amd/display/dc/link/link_dpms.c | 13 +++++++++++-- 1 file changed, 11 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/amd/display/dc/link/link_dpms.c b/drivers/gpu/drm/amd/display/dc/link/link_dpms.c index 268626e73c54..53c961f86d43 100644 --- a/drivers/gpu/drm/amd/display/dc/link/link_dpms.c +++ b/drivers/gpu/drm/amd/display/dc/link/link_dpms.c @@ -148,6 +148,7 @@ void link_blank_dp_stream(struct dc_link *link, bool hw_init) void link_set_all_streams_dpms_off_for_link(struct dc_link *link) { struct pipe_ctx *pipes[MAX_PIPES]; + struct dc_stream_state *streams[MAX_PIPES]; struct dc_state *state = link->dc->current_state; uint8_t count; int i; @@ -160,10 +161,18 @@ void link_set_all_streams_dpms_off_for_link(struct dc_link *link) link_get_master_pipes_with_dpms_on(link, state, &count, pipes); + /* The subsequent call to dc_commit_updates_for_stream for a full update + * will release the current state and swap to a new state. Releasing the + * current state results in the stream pointers in the pipe_ctx structs + * to be zero'd. Hence, cache all streams prior to dc_commit_updates_for_stream. + */ + for (i = 0; i < count; i++) + streams[i] = pipes[i]->stream; + for (i = 0; i < count; i++) { - stream_update.stream = pipes[i]->stream; + stream_update.stream = streams[i]; dc_commit_updates_for_stream(link->ctx->dc, NULL, 0, - pipes[i]->stream, &stream_update, + streams[i], &stream_update, state); } From 2ddac70fed50485aa4ae49cdb7478ce41d8d4715 Mon Sep 17 00:00:00 2001 From: Aurabindo Pillai Date: Tue, 29 Apr 2025 05:03:30 +0800 Subject: [PATCH 10/24] drm/amd/display: check stream id dml21 wrapper to get plane_id [Why & How] Fix a false positive warning which occurs due to lack of correct checks when querying plane_id in DML21. This fixes the warning when performing a mode1 reset (cat /sys/kernel/debug/dri/1/amdgpu_gpu_recover): [ 35.751250] WARNING: CPU: 11 PID: 326 at /tmp/amd.PHpyAl7v/amd/amdgpu/../display/dc/dml2/dml2_dc_resource_mgmt.c:91 dml2_map_dc_pipes+0x243d/0x3f40 [amdgpu] [ 35.751434] Modules linked in: amdgpu(OE) amddrm_ttm_helper(OE) amdttm(OE) amddrm_buddy(OE) amdxcp(OE) amddrm_exec(OE) amd_sched(OE) amdkcl(OE) drm_suballoc_helper drm_ttm_helper ttm drm_display_helper cec rc_core i2c_algo_bit rfcomm qrtr cmac algif_hash algif_skcipher af_alg bnep amd_atl intel_rapl_msr intel_rapl_common snd_hda_codec_hdmi snd_hda_intel edac_mce_amd snd_intel_dspcfg snd_intel_sdw_acpi snd_hda_codec kvm_amd snd_hda_core snd_hwdep snd_pcm kvm snd_seq_midi snd_seq_midi_event snd_rawmidi crct10dif_pclmul polyval_clmulni polyval_generic btusb ghash_clmulni_intel sha256_ssse3 btrtl sha1_ssse3 snd_seq btintel aesni_intel btbcm btmtk snd_seq_device crypto_simd sunrpc cryptd bluetooth snd_timer ccp binfmt_misc rapl snd i2c_piix4 wmi_bmof gigabyte_wmi k10temp i2c_smbus soundcore gpio_amdpt mac_hid sch_fq_codel msr parport_pc ppdev lp parport efi_pstore nfnetlink dmi_sysfs ip_tables x_tables autofs4 hid_generic usbhid hid crc32_pclmul igc ahci xhci_pci libahci xhci_pci_renesas video wmi [ 35.751501] CPU: 11 UID: 0 PID: 326 Comm: kworker/u64:9 Tainted: G OE 6.11.0-21-generic #21~24.04.1-Ubuntu [ 35.751504] Tainted: [O]=OOT_MODULE, [E]=UNSIGNED_MODULE [ 35.751505] Hardware name: Gigabyte Technology Co., Ltd. X670E AORUS PRO X/X670E AORUS PRO X, BIOS F30 05/22/2024 [ 35.751506] Workqueue: amdgpu-reset-dev amdgpu_debugfs_reset_work [amdgpu] [ 35.751638] RIP: 0010:dml2_map_dc_pipes+0x243d/0x3f40 [amdgpu] [ 35.751794] Code: 6d 0c 00 00 8b 84 24 88 00 00 00 41 3b 44 9c 20 0f 84 fc 07 00 00 48 83 c3 01 48 83 fb 06 75 b3 4c 8b 64 24 68 4c 8b 6c 24 40 <0f> 0b b8 06 00 00 00 49 8b 94 24 a0 49 00 00 89 c3 83 f8 07 0f 87 [ 35.751796] RSP: 0018:ffffbfa3805d7680 EFLAGS: 00010246 [ 35.751798] RAX: 0000000000010000 RBX: 0000000000000006 RCX: 0000000000000000 [ 35.751799] RDX: 0000000000000000 RSI: 0000000000000005 RDI: 0000000000000000 [ 35.751800] RBP: ffffbfa3805d78f0 R08: 0000000000000000 R09: 0000000000000000 [ 35.751801] R10: 0000000000000000 R11: 0000000000000000 R12: ffffbfa383249000 [ 35.751802] R13: ffffa0e68f280000 R14: ffffbfa383249658 R15: 0000000000000000 [ 35.751803] FS: 0000000000000000(0000) GS:ffffa0edbe580000(0000) knlGS:0000000000000000 [ 35.751804] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 35.751805] CR2: 00005d847ef96c58 CR3: 000000041de3e000 CR4: 0000000000f50ef0 [ 35.751806] PKRU: 55555554 [ 35.751807] Call Trace: [ 35.751810] [ 35.751816] ? show_regs+0x6c/0x80 [ 35.751820] ? __warn+0x88/0x140 [ 35.751822] ? dml2_map_dc_pipes+0x243d/0x3f40 [amdgpu] [ 35.751964] ? report_bug+0x182/0x1b0 [ 35.751969] ? handle_bug+0x6e/0xb0 [ 35.751972] ? exc_invalid_op+0x18/0x80 [ 35.751974] ? asm_exc_invalid_op+0x1b/0x20 [ 35.751978] ? dml2_map_dc_pipes+0x243d/0x3f40 [amdgpu] [ 35.752117] ? math_pow+0x48/0xa0 [amdgpu] [ 35.752256] ? srso_alias_return_thunk+0x5/0xfbef5 [ 35.752260] ? math_pow+0x48/0xa0 [amdgpu] [ 35.752400] ? srso_alias_return_thunk+0x5/0xfbef5 [ 35.752403] ? math_pow+0x11/0xa0 [amdgpu] [ 35.752524] ? srso_alias_return_thunk+0x5/0xfbef5 [ 35.752526] ? core_dcn4_mode_programming+0xe4d/0x20d0 [amdgpu] [ 35.752663] ? srso_alias_return_thunk+0x5/0xfbef5 [ 35.752669] dml21_validate+0x3d4/0x980 [amdgpu] Reviewed-by: Austin Zheng Signed-off-by: Aurabindo Pillai Signed-off-by: Ray Wu Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher (cherry picked from commit f8ad62c0a93e5dd94243e10f1b742232e4d6411e) --- .../dc/dml2/dml21/dml21_translation_helper.c | 20 ++++++++++--------- 1 file changed, 11 insertions(+), 9 deletions(-) diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_translation_helper.c b/drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_translation_helper.c index 0c8ec30ea672..731fbd4bc600 100644 --- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_translation_helper.c +++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_translation_helper.c @@ -910,7 +910,7 @@ static void populate_dml21_plane_config_from_plane_state(struct dml2_context *dm } //TODO : Could be possibly moved to a common helper layer. -static bool dml21_wrapper_get_plane_id(const struct dc_state *context, const struct dc_plane_state *plane, unsigned int *plane_id) +static bool dml21_wrapper_get_plane_id(const struct dc_state *context, unsigned int stream_id, const struct dc_plane_state *plane, unsigned int *plane_id) { int i, j; @@ -918,10 +918,12 @@ static bool dml21_wrapper_get_plane_id(const struct dc_state *context, const str return false; for (i = 0; i < context->stream_count; i++) { - for (j = 0; j < context->stream_status[i].plane_count; j++) { - if (context->stream_status[i].plane_states[j] == plane) { - *plane_id = (i << 16) | j; - return true; + if (context->streams[i]->stream_id == stream_id) { + for (j = 0; j < context->stream_status[i].plane_count; j++) { + if (context->stream_status[i].plane_states[j] == plane) { + *plane_id = (i << 16) | j; + return true; + } } } } @@ -944,14 +946,14 @@ static unsigned int map_stream_to_dml21_display_cfg(const struct dml2_context *d return location; } -static unsigned int map_plane_to_dml21_display_cfg(const struct dml2_context *dml_ctx, +static unsigned int map_plane_to_dml21_display_cfg(const struct dml2_context *dml_ctx, unsigned int stream_id, const struct dc_plane_state *plane, const struct dc_state *context) { unsigned int plane_id; int i = 0; int location = -1; - if (!dml21_wrapper_get_plane_id(context, plane, &plane_id)) { + if (!dml21_wrapper_get_plane_id(context, stream_id, plane, &plane_id)) { ASSERT(false); return -1; } @@ -1037,7 +1039,7 @@ bool dml21_map_dc_state_into_dml_display_cfg(const struct dc *in_dc, struct dc_s dml_dispcfg->plane_descriptors[disp_cfg_plane_location].stream_index = disp_cfg_stream_location; } else { for (plane_index = 0; plane_index < context->stream_status[stream_index].plane_count; plane_index++) { - disp_cfg_plane_location = map_plane_to_dml21_display_cfg(dml_ctx, context->stream_status[stream_index].plane_states[plane_index], context); + disp_cfg_plane_location = map_plane_to_dml21_display_cfg(dml_ctx, context->streams[stream_index]->stream_id, context->stream_status[stream_index].plane_states[plane_index], context); if (disp_cfg_plane_location < 0) disp_cfg_plane_location = dml_dispcfg->num_planes++; @@ -1048,7 +1050,7 @@ bool dml21_map_dc_state_into_dml_display_cfg(const struct dc *in_dc, struct dc_s populate_dml21_plane_config_from_plane_state(dml_ctx, &dml_dispcfg->plane_descriptors[disp_cfg_plane_location], context->stream_status[stream_index].plane_states[plane_index], context, stream_index); dml_dispcfg->plane_descriptors[disp_cfg_plane_location].stream_index = disp_cfg_stream_location; - if (dml21_wrapper_get_plane_id(context, context->stream_status[stream_index].plane_states[plane_index], &dml_ctx->v21.dml_to_dc_pipe_mapping.disp_cfg_to_plane_id[disp_cfg_plane_location])) + if (dml21_wrapper_get_plane_id(context, context->streams[stream_index]->stream_id, context->stream_status[stream_index].plane_states[plane_index], &dml_ctx->v21.dml_to_dc_pipe_mapping.disp_cfg_to_plane_id[disp_cfg_plane_location])) dml_ctx->v21.dml_to_dc_pipe_mapping.disp_cfg_to_plane_id_valid[disp_cfg_plane_location] = true; /* apply forced pstate policy */ From a3b7e65b6be59e686e163fa1ceb0922f996897c2 Mon Sep 17 00:00:00 2001 From: Melissa Wen Date: Wed, 30 Apr 2025 11:11:47 -0300 Subject: [PATCH 11/24] drm/amd/display: Fix null check of pipe_ctx->plane_state for update_dchubp_dpp Similar to commit 6a057072ddd1 ("drm/amd/display: Fix null check for pipe_ctx->plane_state in dcn20_program_pipe") that addresses a null pointer dereference on dcn20_update_dchubp_dpp. This is the same function hooked for update_dchubp_dpp in dcn401, with the same issue. Fix possible null pointer deference on dcn401_program_pipe too. Fixes: 63ab80d9ac0a ("drm/amd/display: DML2.1 Post-Si Cleanup") Signed-off-by: Melissa Wen Reviewed-by: Alex Hung Signed-off-by: Alex Deucher (cherry picked from commit d8d47f739752227957d8efc0cb894761bfe1d879) --- drivers/gpu/drm/amd/display/dc/hwss/dcn401/dcn401_hwseq.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn401/dcn401_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn401/dcn401_hwseq.c index 5489f3d431f6..3af6a3402b89 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn401/dcn401_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn401/dcn401_hwseq.c @@ -1980,9 +1980,9 @@ void dcn401_program_pipe( dc->res_pool->hubbub, pipe_ctx->plane_res.hubp->inst, pipe_ctx->hubp_regs.det_size); } - if (pipe_ctx->update_flags.raw || - (pipe_ctx->plane_state && pipe_ctx->plane_state->update_flags.raw) || - pipe_ctx->stream->update_flags.raw) + if (pipe_ctx->plane_state && (pipe_ctx->update_flags.raw || + pipe_ctx->plane_state->update_flags.raw || + pipe_ctx->stream->update_flags.raw)) dc->hwss.update_dchubp_dpp(dc, pipe_ctx, context); if (pipe_ctx->plane_state && (pipe_ctx->update_flags.bits.enable || From d33724ffb743d3d2698bd969e29253ae0cff9739 Mon Sep 17 00:00:00 2001 From: Wayne Lin Date: Tue, 13 May 2025 11:20:24 +0800 Subject: [PATCH 12/24] drm/amd/display: Avoid flooding unnecessary info messages It's expected that we'll encounter temporary exceptions during aux transactions. Adjust logging from drm_info to drm_dbg_dp to prevent flooding with unnecessary log messages. Fixes: 3637e457eb00 ("drm/amd/display: Fix wrong handling for AUX_DEFER case") Cc: Mario Limonciello Cc: Alex Deucher Signed-off-by: Wayne Lin Acked-by: Alex Deucher Link: https://lore.kernel.org/r/20250513032026.838036-1-Wayne.Lin@amd.com Signed-off-by: Mario Limonciello Signed-off-by: Alex Deucher (cherry picked from commit 9a9c3e1fe5256da14a0a307dff0478f90c55fc8c) Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c index c93eef6f7a6b..5cdbc86ef8f5 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c @@ -107,7 +107,7 @@ static ssize_t dm_dp_aux_transfer(struct drm_dp_aux *aux, if (payload.write && result >= 0) { if (result) { /*one byte indicating partially written bytes*/ - drm_info(adev_to_drm(adev), "amdgpu: AUX partially written\n"); + drm_dbg_dp(adev_to_drm(adev), "amdgpu: AUX partially written\n"); result = payload.data[0]; } else if (!payload.reply[0]) /*I2C_ACK|AUX_ACK*/ @@ -133,11 +133,11 @@ static ssize_t dm_dp_aux_transfer(struct drm_dp_aux *aux, break; } - drm_info(adev_to_drm(adev), "amdgpu: DP AUX transfer fail:%d\n", operation_result); + drm_dbg_dp(adev_to_drm(adev), "amdgpu: DP AUX transfer fail:%d\n", operation_result); } if (payload.reply[0]) - drm_info(adev_to_drm(adev), "amdgpu: AUX reply command not ACK: 0x%02x.", + drm_dbg_dp(adev_to_drm(adev), "amdgpu: AUX reply command not ACK: 0x%02x.", payload.reply[0]); return result; From 2d2f82e1a8a118d26c2489674b6f6cbb41d04376 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Thomas=20Hellstr=C3=B6m?= Date: Thu, 8 May 2025 13:29:31 +0200 Subject: [PATCH 13/24] drm/xe: Fix the gem shrinker name MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The xe buffer object shrinker name is visible in the /shrinker directory and most if not all other shinkers follow a naming convention that looks like -_: Follow the same convention for xe, changing the name to drm-xe_gem:. Other shrinkers typically use the device node for but since drm drivers typically don't have a single unique device- node, instead use the unique name in the drm device. Fixes: 00c8efc3180f ("drm/xe: Add a shrinker for xe bos") Cc: Matthew Brost Signed-off-by: Thomas Hellström Reviewed-by: Francois Dugast Link: https://lore.kernel.org/r/20250508112931.3347-1-thomas.hellstrom@linux.intel.com (cherry picked from commit 243bf99e2fe75edf8df1711c1377b6fc020b806c) Signed-off-by: Lucas De Marchi --- drivers/gpu/drm/xe/xe_shrinker.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/xe/xe_shrinker.c b/drivers/gpu/drm/xe/xe_shrinker.c index 8184390f9c7b..86d47aaf0358 100644 --- a/drivers/gpu/drm/xe/xe_shrinker.c +++ b/drivers/gpu/drm/xe/xe_shrinker.c @@ -227,7 +227,7 @@ struct xe_shrinker *xe_shrinker_create(struct xe_device *xe) if (!shrinker) return ERR_PTR(-ENOMEM); - shrinker->shrink = shrinker_alloc(0, "xe system shrinker"); + shrinker->shrink = shrinker_alloc(0, "drm-xe_gem:%s", xe->drm.unique); if (!shrinker->shrink) { kfree(shrinker); return ERR_PTR(-ENOMEM); From 72c7d62583ebce7baeb61acce6057c361f73be4a Mon Sep 17 00:00:00 2001 From: Hyejeong Choi Date: Mon, 12 May 2025 21:06:38 -0500 Subject: [PATCH 14/24] dma-buf: insert memory barrier before updating num_fences MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit smp_store_mb() inserts memory barrier after storing operation. It is different with what the comment is originally aiming so Null pointer dereference can be happened if memory update is reordered. Signed-off-by: Hyejeong Choi Fixes: a590d0fdbaa5 ("dma-buf: Update reservation shared_count after adding the new fence") CC: stable@vger.kernel.org Reviewed-by: Christian König Link: https://lore.kernel.org/r/20250513020638.GA2329653@au1-maretx-p37.eng.sarc.samsung.com Signed-off-by: Christian König --- drivers/dma-buf/dma-resv.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/drivers/dma-buf/dma-resv.c b/drivers/dma-buf/dma-resv.c index 5f8d010516f0..b1ef4546346d 100644 --- a/drivers/dma-buf/dma-resv.c +++ b/drivers/dma-buf/dma-resv.c @@ -320,8 +320,9 @@ void dma_resv_add_fence(struct dma_resv *obj, struct dma_fence *fence, count++; dma_resv_list_set(fobj, i, fence, usage); - /* pointer update must be visible before we extend the num_fences */ - smp_store_mb(fobj->num_fences, count); + /* fence update must be visible before we extend the num_fences */ + smp_wmb(); + fobj->num_fences = count; } EXPORT_SYMBOL(dma_resv_add_fence); From fe14c0f096f58d2569e587e9f4b05d772272bbb4 Mon Sep 17 00:00:00 2001 From: Melissa Wen Date: Tue, 22 Apr 2025 11:58:11 -0300 Subject: [PATCH 15/24] Revert "drm/amd/display: Hardware cursor changes color when switched to software cursor" MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This reverts commit 272e6aab14bbf98d7a06b2b1cd6308a02d4a10a1. Applying degamma curve to the cursor by default breaks Linux userspace expectation. On Linux, AMD display manager enables cursor degamma ROM just for implict sRGB on HW versions where degamma is split into two blocks: degamma ROM for pre-defined TFs and `gamma correction` for user/custom curves, and degamma ROM settings doesn't apply to cursor plane. Link: https://gitlab.freedesktop.org/drm/amd/-/issues/1513 Link: https://gitlab.freedesktop.org/drm/amd/-/issues/2803 Reported-by: Michel Dänzer Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/4144 Signed-off-by: Melissa Wen Reviewed-by: Alex Hung Signed-off-by: Alex Deucher (cherry picked from commit f6a305d4748801a6c799ae9375b2ecff3aed094b) Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/display/dc/dpp/dcn401/dcn401_dpp_cm.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/amd/display/dc/dpp/dcn401/dcn401_dpp_cm.c b/drivers/gpu/drm/amd/display/dc/dpp/dcn401/dcn401_dpp_cm.c index 1236e0f9a256..712aff7e17f7 100644 --- a/drivers/gpu/drm/amd/display/dc/dpp/dcn401/dcn401_dpp_cm.c +++ b/drivers/gpu/drm/amd/display/dc/dpp/dcn401/dcn401_dpp_cm.c @@ -120,10 +120,11 @@ void dpp401_set_cursor_attributes( enum dc_cursor_color_format color_format = cursor_attributes->color_format; int cur_rom_en = 0; - // DCN4 should always do Cursor degamma for Cursor Color modes if (color_format == CURSOR_MODE_COLOR_PRE_MULTIPLIED_ALPHA || color_format == CURSOR_MODE_COLOR_UN_PRE_MULTIPLIED_ALPHA) { - cur_rom_en = 1; + if (cursor_attributes->attribute_flags.bits.ENABLE_CURSOR_DEGAMMA) { + cur_rom_en = 1; + } } REG_UPDATE_3(CURSOR0_CONTROL, From ee7360fc27d6045510f8fe459b5649b2af27811a Mon Sep 17 00:00:00 2001 From: "David (Ming Qiang) Wu" Date: Mon, 12 May 2025 15:14:43 -0400 Subject: [PATCH 16/24] drm/amdgpu: read back register after written for VCN v4.0.5 On VCN v4.0.5 there is a race condition where the WPTR is not updated after starting from idle when doorbell is used. Adding register read-back after written at function end is to ensure all register writes are done before they can be used. Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/12528 Signed-off-by: David (Ming Qiang) Wu Reviewed-by: Mario Limonciello Tested-by: Mario Limonciello Reviewed-by: Alex Deucher Reviewed-by: Ruijing Dong Signed-off-by: Alex Deucher (cherry picked from commit 07c9db090b86e5211188e1b351303fbc673378cf) Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/amdgpu/vcn_v4_0_5.c | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_5.c b/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_5.c index a1171e6152ed..f11df9c2ec13 100644 --- a/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_5.c +++ b/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_5.c @@ -1023,6 +1023,10 @@ static int vcn_v4_0_5_start_dpg_mode(struct amdgpu_vcn_inst *vinst, ring->doorbell_index << VCN_RB1_DB_CTRL__OFFSET__SHIFT | VCN_RB1_DB_CTRL__EN_MASK); + /* Keeping one read-back to ensure all register writes are done, otherwise + * it may introduce race conditions */ + RREG32_SOC15(VCN, inst_idx, regVCN_RB1_DB_CTRL); + return 0; } @@ -1205,6 +1209,10 @@ static int vcn_v4_0_5_start(struct amdgpu_vcn_inst *vinst) WREG32_SOC15(VCN, i, regVCN_RB_ENABLE, tmp); fw_shared->sq.queue_mode &= ~(FW_QUEUE_RING_RESET | FW_QUEUE_DPG_HOLD_OFF); + /* Keeping one read-back to ensure all register writes are done, otherwise + * it may introduce race conditions */ + RREG32_SOC15(VCN, i, regVCN_RB_ENABLE); + return 0; } From b1f704107cf27906a9cea542b626b96019104663 Mon Sep 17 00:00:00 2001 From: Aradhya Bhatia Date: Mon, 12 May 2025 06:50:04 +0000 Subject: [PATCH 17/24] drm/xe/xe2hpg: Add Wa_22021007897 Add Wa_22021007897 for the Xe2_HPG (graphics version: 20.01) IP. It is a permanent workaround, and applicable on all the steppings. Reviewed-by: Gustavo Sousa Reviewed-by: Tejas Upadhyay Signed-off-by: Aradhya Bhatia Link: https://lore.kernel.org/r/20250512065004.2576-1-aradhya.bhatia@intel.com Signed-off-by: Matt Roper (cherry picked from commit e5c13e2c505b73a8667ef9a0fd5cbd4227e483e6) Signed-off-by: Lucas De Marchi --- drivers/gpu/drm/xe/regs/xe_gt_regs.h | 1 + drivers/gpu/drm/xe/xe_wa.c | 4 ++++ 2 files changed, 5 insertions(+) diff --git a/drivers/gpu/drm/xe/regs/xe_gt_regs.h b/drivers/gpu/drm/xe/regs/xe_gt_regs.h index da1f198ac107..181913967ac9 100644 --- a/drivers/gpu/drm/xe/regs/xe_gt_regs.h +++ b/drivers/gpu/drm/xe/regs/xe_gt_regs.h @@ -157,6 +157,7 @@ #define XEHPG_SC_INSTDONE_EXTRA2 XE_REG_MCR(0x7108) #define COMMON_SLICE_CHICKEN4 XE_REG(0x7300, XE_REG_OPTION_MASKED) +#define SBE_PUSH_CONSTANT_BEHIND_FIX_ENABLE REG_BIT(12) #define DISABLE_TDC_LOAD_BALANCING_CALC REG_BIT(6) #define COMMON_SLICE_CHICKEN3 XE_REG(0x7304, XE_REG_OPTION_MASKED) diff --git a/drivers/gpu/drm/xe/xe_wa.c b/drivers/gpu/drm/xe/xe_wa.c index 24f644c0a673..2f833f0d575f 100644 --- a/drivers/gpu/drm/xe/xe_wa.c +++ b/drivers/gpu/drm/xe/xe_wa.c @@ -815,6 +815,10 @@ static const struct xe_rtp_entry_sr lrc_was[] = { XE_RTP_RULES(GRAPHICS_VERSION(2001), ENGINE_CLASS(RENDER)), XE_RTP_ACTIONS(SET(CHICKEN_RASTER_1, DIS_CLIP_NEGATIVE_BOUNDING_BOX)) }, + { XE_RTP_NAME("22021007897"), + XE_RTP_RULES(GRAPHICS_VERSION(2001), ENGINE_CLASS(RENDER)), + XE_RTP_ACTIONS(SET(COMMON_SLICE_CHICKEN4, SBE_PUSH_CONSTANT_BEHIND_FIX_ENABLE)) + }, /* Xe3_LPG */ { XE_RTP_NAME("14021490052"), From 7bd68ce21d39150d80806233965326359f517b78 Mon Sep 17 00:00:00 2001 From: Himal Prasad Ghimiray Date: Mon, 12 May 2025 06:54:55 -0700 Subject: [PATCH 18/24] drm/gpusvm: Introduce devmem_only flag for allocation This commit adds a new flag, devmem_only, to the drm_gpusvm structure. The purpose of this flag is to ensure that the get_pages function allocates memory exclusively from the device's memory. If the allocation from device memory fails, the function will return an -EFAULT error. Required for shared CPU and GPU atomics on certain devices. v3: - s/vram_only/devmem_only/ Fixes: 99624bdff867 ("drm/gpusvm: Add support for GPU Shared Virtual Memory") Cc: stable@vger.kernel.org Signed-off-by: Matthew Brost Signed-off-by: Himal Prasad Ghimiray Reviewed-by: Matthew Brost Link: https://lore.kernel.org/r/20250512135500.1405019-2-matthew.brost@intel.com (cherry picked from commit 8a9b978ebd47df9e0694c34748c2d6fa0c31eb4d) Signed-off-by: Lucas De Marchi --- drivers/gpu/drm/drm_gpusvm.c | 5 +++++ include/drm/drm_gpusvm.h | 2 ++ 2 files changed, 7 insertions(+) diff --git a/drivers/gpu/drm/drm_gpusvm.c b/drivers/gpu/drm/drm_gpusvm.c index de424e670995..a58d03e6cac2 100644 --- a/drivers/gpu/drm/drm_gpusvm.c +++ b/drivers/gpu/drm/drm_gpusvm.c @@ -1454,6 +1454,11 @@ map_pages: goto err_unmap; } + if (ctx->devmem_only) { + err = -EFAULT; + goto err_unmap; + } + addr = dma_map_page(gpusvm->drm->dev, page, 0, PAGE_SIZE << order, diff --git a/include/drm/drm_gpusvm.h b/include/drm/drm_gpusvm.h index df120b4d1f83..9fd25fc880a4 100644 --- a/include/drm/drm_gpusvm.h +++ b/include/drm/drm_gpusvm.h @@ -286,6 +286,7 @@ struct drm_gpusvm { * @in_notifier: entering from a MMU notifier * @read_only: operating on read-only memory * @devmem_possible: possible to use device memory + * @devmem_only: use only device memory * * Context that is DRM GPUSVM is operating in (i.e. user arguments). */ @@ -294,6 +295,7 @@ struct drm_gpusvm_ctx { unsigned int in_notifier :1; unsigned int read_only :1; unsigned int devmem_possible :1; + unsigned int devmem_only :1; }; int drm_gpusvm_init(struct drm_gpusvm *gpusvm, From 794f5493f518916380578f14d999de92b930b609 Mon Sep 17 00:00:00 2001 From: Matthew Brost Date: Mon, 12 May 2025 06:54:56 -0700 Subject: [PATCH 19/24] drm/xe: Strict migration policy for atomic SVM faults MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Mixing GPU and CPU atomics does not work unless a strict migration policy of GPU atomics must be device memory. Enforce a policy of must be in VRAM with a retry loop of 3 attempts, if retry loop fails abort fault. Removing always_migrate_to_vram modparam as we now have real migration policy. v2: - Only retry migration on atomics - Drop alway migrate modparam v3: - Only set vram_only on DGFX (Himal) - Bail on get_pages failure if vram_only and retry count exceeded (Himal) - s/vram_only/devmem_only - Update xe_svm_range_is_valid to accept devmem_only argument v4: - Fix logic bug get_pages failure v5: - Fix commit message (Himal) - Mention removing always_migrate_to_vram in commit message (Lucas) - Fix xe_svm_range_is_valid to check for devmem pages - Bail on devmem_only && !migrate_devmem (Thomas) v6: - Add READ_ONCE barriers for opportunistic checks (Thomas) - Pair READ_ONCE with WRITE_ONCE (Thomas) v7: - Adjust comments (Thomas) Fixes: 2f118c949160 ("drm/xe: Add SVM VRAM migration") Cc: stable@vger.kernel.org Signed-off-by: Himal Prasad Ghimiray Signed-off-by: Matthew Brost Acked-by: Himal Prasad Ghimiray Reviewed-by: Thomas Hellström Link: https://lore.kernel.org/r/20250512135500.1405019-3-matthew.brost@intel.com (cherry picked from commit a9ac0fa455b050d03e3032501368048fb284d318) Signed-off-by: Lucas De Marchi --- drivers/gpu/drm/drm_gpusvm.c | 23 +++++-- drivers/gpu/drm/xe/xe_module.c | 3 - drivers/gpu/drm/xe/xe_module.h | 1 - drivers/gpu/drm/xe/xe_pt.c | 14 ++++- drivers/gpu/drm/xe/xe_svm.c | 111 +++++++++++++++++++++++++-------- drivers/gpu/drm/xe/xe_svm.h | 5 -- include/drm/drm_gpusvm.h | 40 +++++++----- 7 files changed, 140 insertions(+), 57 deletions(-) diff --git a/drivers/gpu/drm/drm_gpusvm.c b/drivers/gpu/drm/drm_gpusvm.c index a58d03e6cac2..41f6616bcf76 100644 --- a/drivers/gpu/drm/drm_gpusvm.c +++ b/drivers/gpu/drm/drm_gpusvm.c @@ -1118,6 +1118,10 @@ static void __drm_gpusvm_range_unmap_pages(struct drm_gpusvm *gpusvm, lockdep_assert_held(&gpusvm->notifier_lock); if (range->flags.has_dma_mapping) { + struct drm_gpusvm_range_flags flags = { + .__flags = range->flags.__flags, + }; + for (i = 0, j = 0; i < npages; j++) { struct drm_pagemap_device_addr *addr = &range->dma_addr[j]; @@ -1131,8 +1135,12 @@ static void __drm_gpusvm_range_unmap_pages(struct drm_gpusvm *gpusvm, dev, *addr); i += 1 << addr->order; } - range->flags.has_devmem_pages = false; - range->flags.has_dma_mapping = false; + + /* WRITE_ONCE pairs with READ_ONCE for opportunistic checks */ + flags.has_devmem_pages = false; + flags.has_dma_mapping = false; + WRITE_ONCE(range->flags.__flags, flags.__flags); + range->dpagemap = NULL; } } @@ -1334,6 +1342,7 @@ int drm_gpusvm_range_get_pages(struct drm_gpusvm *gpusvm, int err = 0; struct dev_pagemap *pagemap; struct drm_pagemap *dpagemap; + struct drm_gpusvm_range_flags flags; retry: hmm_range.notifier_seq = mmu_interval_read_begin(notifier); @@ -1378,7 +1387,8 @@ map_pages: */ drm_gpusvm_notifier_lock(gpusvm); - if (range->flags.unmapped) { + flags.__flags = range->flags.__flags; + if (flags.unmapped) { drm_gpusvm_notifier_unlock(gpusvm); err = -EFAULT; goto err_free; @@ -1474,14 +1484,17 @@ map_pages: } i += 1 << order; num_dma_mapped = i; - range->flags.has_dma_mapping = true; + flags.has_dma_mapping = true; } if (zdd) { - range->flags.has_devmem_pages = true; + flags.has_devmem_pages = true; range->dpagemap = dpagemap; } + /* WRITE_ONCE pairs with READ_ONCE for opportunistic checks */ + WRITE_ONCE(range->flags.__flags, flags.__flags); + drm_gpusvm_notifier_unlock(gpusvm); kvfree(pfns); set_seqno: diff --git a/drivers/gpu/drm/xe/xe_module.c b/drivers/gpu/drm/xe/xe_module.c index 9f4632e39a1a..e861c694f336 100644 --- a/drivers/gpu/drm/xe/xe_module.c +++ b/drivers/gpu/drm/xe/xe_module.c @@ -29,9 +29,6 @@ struct xe_modparam xe_modparam = { module_param_named(svm_notifier_size, xe_modparam.svm_notifier_size, uint, 0600); MODULE_PARM_DESC(svm_notifier_size, "Set the svm notifier size(in MiB), must be power of 2"); -module_param_named(always_migrate_to_vram, xe_modparam.always_migrate_to_vram, bool, 0444); -MODULE_PARM_DESC(always_migrate_to_vram, "Always migrate to VRAM on GPU fault"); - module_param_named_unsafe(force_execlist, xe_modparam.force_execlist, bool, 0444); MODULE_PARM_DESC(force_execlist, "Force Execlist submission"); diff --git a/drivers/gpu/drm/xe/xe_module.h b/drivers/gpu/drm/xe/xe_module.h index 84339e509c80..5a3bfea8b7b4 100644 --- a/drivers/gpu/drm/xe/xe_module.h +++ b/drivers/gpu/drm/xe/xe_module.h @@ -12,7 +12,6 @@ struct xe_modparam { bool force_execlist; bool probe_display; - bool always_migrate_to_vram; u32 force_vram_bar_size; int guc_log_level; char *guc_firmware_path; diff --git a/drivers/gpu/drm/xe/xe_pt.c b/drivers/gpu/drm/xe/xe_pt.c index ffaf0d02dc7d..856038553b81 100644 --- a/drivers/gpu/drm/xe/xe_pt.c +++ b/drivers/gpu/drm/xe/xe_pt.c @@ -2232,11 +2232,19 @@ static void op_commit(struct xe_vm *vm, } case DRM_GPUVA_OP_DRIVER: { + /* WRITE_ONCE pairs with READ_ONCE in xe_svm.c */ + if (op->subop == XE_VMA_SUBOP_MAP_RANGE) { - op->map_range.range->tile_present |= BIT(tile->id); - op->map_range.range->tile_invalidated &= ~BIT(tile->id); + WRITE_ONCE(op->map_range.range->tile_present, + op->map_range.range->tile_present | + BIT(tile->id)); + WRITE_ONCE(op->map_range.range->tile_invalidated, + op->map_range.range->tile_invalidated & + ~BIT(tile->id)); } else if (op->subop == XE_VMA_SUBOP_UNMAP_RANGE) { - op->unmap_range.range->tile_present &= ~BIT(tile->id); + WRITE_ONCE(op->unmap_range.range->tile_present, + op->unmap_range.range->tile_present & + ~BIT(tile->id)); } break; } diff --git a/drivers/gpu/drm/xe/xe_svm.c b/drivers/gpu/drm/xe/xe_svm.c index 24c578e1170e..e65d3b820ee0 100644 --- a/drivers/gpu/drm/xe/xe_svm.c +++ b/drivers/gpu/drm/xe/xe_svm.c @@ -15,8 +15,17 @@ static bool xe_svm_range_in_vram(struct xe_svm_range *range) { - /* Not reliable without notifier lock */ - return range->base.flags.has_devmem_pages; + /* + * Advisory only check whether the range is currently backed by VRAM + * memory. + */ + + struct drm_gpusvm_range_flags flags = { + /* Pairs with WRITE_ONCE in drm_gpusvm.c */ + .__flags = READ_ONCE(range->base.flags.__flags), + }; + + return flags.has_devmem_pages; } static bool xe_svm_range_has_vram_binding(struct xe_svm_range *range) @@ -645,9 +654,16 @@ void xe_svm_fini(struct xe_vm *vm) } static bool xe_svm_range_is_valid(struct xe_svm_range *range, - struct xe_tile *tile) + struct xe_tile *tile, + bool devmem_only) { - return (range->tile_present & ~range->tile_invalidated) & BIT(tile->id); + /* + * Advisory only check whether the range currently has a valid mapping, + * READ_ONCE pairs with WRITE_ONCE in xe_pt.c + */ + return ((READ_ONCE(range->tile_present) & + ~READ_ONCE(range->tile_invalidated)) & BIT(tile->id)) && + (!devmem_only || xe_svm_range_in_vram(range)); } static struct xe_vram_region *tile_to_vr(struct xe_tile *tile) @@ -712,6 +728,36 @@ unlock: return err; } +static bool supports_4K_migration(struct xe_device *xe) +{ + if (xe->info.vram_flags & XE_VRAM_FLAGS_NEED64K) + return false; + + return true; +} + +static bool xe_svm_range_needs_migrate_to_vram(struct xe_svm_range *range, + struct xe_vma *vma) +{ + struct xe_vm *vm = range_to_vm(&range->base); + u64 range_size = xe_svm_range_size(range); + + if (!range->base.flags.migrate_devmem) + return false; + + if (xe_svm_range_in_vram(range)) { + drm_dbg(&vm->xe->drm, "Range is already in VRAM\n"); + return false; + } + + if (range_size <= SZ_64K && !supports_4K_migration(vm->xe)) { + drm_dbg(&vm->xe->drm, "Platform doesn't support SZ_4K range migration\n"); + return false; + } + + return true; +} + /** * xe_svm_handle_pagefault() - SVM handle page fault * @vm: The VM. @@ -735,11 +781,14 @@ int xe_svm_handle_pagefault(struct xe_vm *vm, struct xe_vma *vma, IS_ENABLED(CONFIG_DRM_XE_DEVMEM_MIRROR), .check_pages_threshold = IS_DGFX(vm->xe) && IS_ENABLED(CONFIG_DRM_XE_DEVMEM_MIRROR) ? SZ_64K : 0, + .devmem_only = atomic && IS_DGFX(vm->xe) && + IS_ENABLED(CONFIG_DRM_XE_DEVMEM_MIRROR), }; struct xe_svm_range *range; struct drm_gpusvm_range *r; struct drm_exec exec; struct dma_fence *fence; + int migrate_try_count = ctx.devmem_only ? 3 : 1; ktime_t end = 0; int err; @@ -758,24 +807,30 @@ retry: if (IS_ERR(r)) return PTR_ERR(r); + if (ctx.devmem_only && !r->flags.migrate_devmem) + return -EACCES; + range = to_xe_range(r); - if (xe_svm_range_is_valid(range, tile)) + if (xe_svm_range_is_valid(range, tile, ctx.devmem_only)) return 0; range_debug(range, "PAGE FAULT"); - /* XXX: Add migration policy, for now migrate range once */ - if (!range->skip_migrate && range->base.flags.migrate_devmem && - xe_svm_range_size(range) >= SZ_64K) { - range->skip_migrate = true; - + if (--migrate_try_count >= 0 && + xe_svm_range_needs_migrate_to_vram(range, vma)) { err = xe_svm_alloc_vram(vm, tile, range, &ctx); if (err) { - drm_dbg(&vm->xe->drm, - "VRAM allocation failed, falling back to " - "retrying fault, asid=%u, errno=%pe\n", - vm->usm.asid, ERR_PTR(err)); - goto retry; + if (migrate_try_count || !ctx.devmem_only) { + drm_dbg(&vm->xe->drm, + "VRAM allocation failed, falling back to retrying fault, asid=%u, errno=%pe\n", + vm->usm.asid, ERR_PTR(err)); + goto retry; + } else { + drm_err(&vm->xe->drm, + "VRAM allocation failed, retry count exceeded, asid=%u, errno=%pe\n", + vm->usm.asid, ERR_PTR(err)); + return err; + } } } @@ -783,15 +838,22 @@ retry: err = drm_gpusvm_range_get_pages(&vm->svm.gpusvm, r, &ctx); /* Corner where CPU mappings have changed */ if (err == -EOPNOTSUPP || err == -EFAULT || err == -EPERM) { - if (err == -EOPNOTSUPP) { - range_debug(range, "PAGE FAULT - EVICT PAGES"); - drm_gpusvm_range_evict(&vm->svm.gpusvm, &range->base); + if (migrate_try_count > 0 || !ctx.devmem_only) { + if (err == -EOPNOTSUPP) { + range_debug(range, "PAGE FAULT - EVICT PAGES"); + drm_gpusvm_range_evict(&vm->svm.gpusvm, + &range->base); + } + drm_dbg(&vm->xe->drm, + "Get pages failed, falling back to retrying, asid=%u, gpusvm=%p, errno=%pe\n", + vm->usm.asid, &vm->svm.gpusvm, ERR_PTR(err)); + range_debug(range, "PAGE FAULT - RETRY PAGES"); + goto retry; + } else { + drm_err(&vm->xe->drm, + "Get pages failed, retry count exceeded, asid=%u, gpusvm=%p, errno=%pe\n", + vm->usm.asid, &vm->svm.gpusvm, ERR_PTR(err)); } - drm_dbg(&vm->xe->drm, - "Get pages failed, falling back to retrying, asid=%u, gpusvm=%p, errno=%pe\n", - vm->usm.asid, &vm->svm.gpusvm, ERR_PTR(err)); - range_debug(range, "PAGE FAULT - RETRY PAGES"); - goto retry; } if (err) { range_debug(range, "PAGE FAULT - FAIL PAGE COLLECT"); @@ -825,9 +887,6 @@ retry_bind: } drm_exec_fini(&exec); - if (xe_modparam.always_migrate_to_vram) - range->skip_migrate = false; - dma_fence_wait(fence, false); dma_fence_put(fence); diff --git a/drivers/gpu/drm/xe/xe_svm.h b/drivers/gpu/drm/xe/xe_svm.h index be306fe7aaa4..fe58ac2f4baa 100644 --- a/drivers/gpu/drm/xe/xe_svm.h +++ b/drivers/gpu/drm/xe/xe_svm.h @@ -36,11 +36,6 @@ struct xe_svm_range { * range. Protected by GPU SVM notifier lock. */ u8 tile_invalidated; - /** - * @skip_migrate: Skip migration to VRAM, protected by GPU fault handler - * locking. - */ - u8 skip_migrate :1; }; #if IS_ENABLED(CONFIG_DRM_GPUSVM) diff --git a/include/drm/drm_gpusvm.h b/include/drm/drm_gpusvm.h index 9fd25fc880a4..653d48dbe1c1 100644 --- a/include/drm/drm_gpusvm.h +++ b/include/drm/drm_gpusvm.h @@ -185,6 +185,31 @@ struct drm_gpusvm_notifier { } flags; }; +/** + * struct drm_gpusvm_range_flags - Structure representing a GPU SVM range flags + * + * @migrate_devmem: Flag indicating whether the range can be migrated to device memory + * @unmapped: Flag indicating if the range has been unmapped + * @partial_unmap: Flag indicating if the range has been partially unmapped + * @has_devmem_pages: Flag indicating if the range has devmem pages + * @has_dma_mapping: Flag indicating if the range has a DMA mapping + * @__flags: Flags for range in u16 form (used for READ_ONCE) + */ +struct drm_gpusvm_range_flags { + union { + struct { + /* All flags below must be set upon creation */ + u16 migrate_devmem : 1; + /* All flags below must be set / cleared under notifier lock */ + u16 unmapped : 1; + u16 partial_unmap : 1; + u16 has_devmem_pages : 1; + u16 has_dma_mapping : 1; + }; + u16 __flags; + }; +}; + /** * struct drm_gpusvm_range - Structure representing a GPU SVM range * @@ -198,11 +223,6 @@ struct drm_gpusvm_notifier { * @dpagemap: The struct drm_pagemap of the device pages we're dma-mapping. * Note this is assuming only one drm_pagemap per range is allowed. * @flags: Flags for range - * @flags.migrate_devmem: Flag indicating whether the range can be migrated to device memory - * @flags.unmapped: Flag indicating if the range has been unmapped - * @flags.partial_unmap: Flag indicating if the range has been partially unmapped - * @flags.has_devmem_pages: Flag indicating if the range has devmem pages - * @flags.has_dma_mapping: Flag indicating if the range has a DMA mapping * * This structure represents a GPU SVM range used for tracking memory ranges * mapped in a DRM device. @@ -216,15 +236,7 @@ struct drm_gpusvm_range { unsigned long notifier_seq; struct drm_pagemap_device_addr *dma_addr; struct drm_pagemap *dpagemap; - struct { - /* All flags below must be set upon creation */ - u16 migrate_devmem : 1; - /* All flags below must be set / cleared under notifier lock */ - u16 unmapped : 1; - u16 partial_unmap : 1; - u16 has_devmem_pages : 1; - u16 has_dma_mapping : 1; - } flags; + struct drm_gpusvm_range_flags flags; }; /** From df8c37810b5a4a5c8bff1f880f59163c800a4985 Mon Sep 17 00:00:00 2001 From: Matthew Brost Date: Mon, 12 May 2025 06:54:57 -0700 Subject: [PATCH 20/24] drm/gpusvm: Add timeslicing support to GPU SVM Add timeslicing support to GPU SVM which will guarantee the GPU a minimum execution time on piece of physical memory before migration back to CPU. Intended to implement strict migration policies which require memory to be in a certain placement for correct execution. Required for shared CPU and GPU atomics on certain devices. Fixes: 99624bdff867 ("drm/gpusvm: Add support for GPU Shared Virtual Memory") Cc: stable@vger.kernel.org Signed-off-by: Matthew Brost Reviewed-by: Himal Prasad Ghimiray Link: https://lore.kernel.org/r/20250512135500.1405019-4-matthew.brost@intel.com (cherry picked from commit 8dc1812b5b3a42311d28eb385eed88e2053ad3cb) Signed-off-by: Lucas De Marchi --- drivers/gpu/drm/drm_gpusvm.c | 9 +++++++++ include/drm/drm_gpusvm.h | 5 +++++ 2 files changed, 14 insertions(+) diff --git a/drivers/gpu/drm/drm_gpusvm.c b/drivers/gpu/drm/drm_gpusvm.c index 41f6616bcf76..4b2f32889f00 100644 --- a/drivers/gpu/drm/drm_gpusvm.c +++ b/drivers/gpu/drm/drm_gpusvm.c @@ -1783,6 +1783,8 @@ int drm_gpusvm_migrate_to_devmem(struct drm_gpusvm *gpusvm, goto err_finalize; /* Upon success bind devmem allocation to range and zdd */ + devmem_allocation->timeslice_expiration = get_jiffies_64() + + msecs_to_jiffies(ctx->timeslice_ms); zdd->devmem_allocation = devmem_allocation; /* Owns ref */ err_finalize: @@ -2003,6 +2005,13 @@ static int __drm_gpusvm_migrate_to_ram(struct vm_area_struct *vas, void *buf; int i, err = 0; + if (page) { + zdd = page->zone_device_data; + if (time_before64(get_jiffies_64(), + zdd->devmem_allocation->timeslice_expiration)) + return 0; + } + start = ALIGN_DOWN(fault_addr, size); end = ALIGN(fault_addr + 1, size); diff --git a/include/drm/drm_gpusvm.h b/include/drm/drm_gpusvm.h index 653d48dbe1c1..eaf704d3d05e 100644 --- a/include/drm/drm_gpusvm.h +++ b/include/drm/drm_gpusvm.h @@ -89,6 +89,7 @@ struct drm_gpusvm_devmem_ops { * @ops: Pointer to the operations structure for GPU SVM device memory * @dpagemap: The struct drm_pagemap of the pages this allocation belongs to. * @size: Size of device memory allocation + * @timeslice_expiration: Timeslice expiration in jiffies */ struct drm_gpusvm_devmem { struct device *dev; @@ -97,6 +98,7 @@ struct drm_gpusvm_devmem { const struct drm_gpusvm_devmem_ops *ops; struct drm_pagemap *dpagemap; size_t size; + u64 timeslice_expiration; }; /** @@ -295,6 +297,8 @@ struct drm_gpusvm { * @check_pages_threshold: Check CPU pages for present if chunk is less than or * equal to threshold. If not present, reduce chunk * size. + * @timeslice_ms: The timeslice MS which in minimum time a piece of memory + * remains with either exclusive GPU or CPU access. * @in_notifier: entering from a MMU notifier * @read_only: operating on read-only memory * @devmem_possible: possible to use device memory @@ -304,6 +308,7 @@ struct drm_gpusvm { */ struct drm_gpusvm_ctx { unsigned long check_pages_threshold; + unsigned long timeslice_ms; unsigned int in_notifier :1; unsigned int read_only :1; unsigned int devmem_possible :1; From 1b36ea2fc6879fed02b675e86867bc422b76f50e Mon Sep 17 00:00:00 2001 From: Matthew Brost Date: Mon, 12 May 2025 06:54:58 -0700 Subject: [PATCH 21/24] drm/xe: Timeslice GPU on atomic SVM fault Ensure GPU can make forward progress on an atomic SVM GPU fault by giving the GPU a timeslice of 5ms v2: - Reduce timeslice to 5ms - Double timeslice on retry - Split out GPU SVM changes into independent patch v5: - Double timeslice in a few more places Fixes: 2f118c949160 ("drm/xe: Add SVM VRAM migration") Cc: stable@vger.kernel.org Signed-off-by: Matthew Brost Reviewed-by: Himal Prasad Ghimiray Link: https://lore.kernel.org/r/20250512135500.1405019-5-matthew.brost@intel.com (cherry picked from commit a5d8d3be1dea8154edbbea481081469627665659) Signed-off-by: Lucas De Marchi --- drivers/gpu/drm/xe/xe_svm.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/drivers/gpu/drm/xe/xe_svm.c b/drivers/gpu/drm/xe/xe_svm.c index e65d3b820ee0..975094c1a582 100644 --- a/drivers/gpu/drm/xe/xe_svm.c +++ b/drivers/gpu/drm/xe/xe_svm.c @@ -783,6 +783,8 @@ int xe_svm_handle_pagefault(struct xe_vm *vm, struct xe_vma *vma, IS_ENABLED(CONFIG_DRM_XE_DEVMEM_MIRROR) ? SZ_64K : 0, .devmem_only = atomic && IS_DGFX(vm->xe) && IS_ENABLED(CONFIG_DRM_XE_DEVMEM_MIRROR), + .timeslice_ms = atomic && IS_DGFX(vm->xe) && + IS_ENABLED(CONFIG_DRM_XE_DEVMEM_MIRROR) ? 5 : 0, }; struct xe_svm_range *range; struct drm_gpusvm_range *r; @@ -819,6 +821,7 @@ retry: if (--migrate_try_count >= 0 && xe_svm_range_needs_migrate_to_vram(range, vma)) { err = xe_svm_alloc_vram(vm, tile, range, &ctx); + ctx.timeslice_ms <<= 1; /* Double timeslice if we have to retry */ if (err) { if (migrate_try_count || !ctx.devmem_only) { drm_dbg(&vm->xe->drm, @@ -838,6 +841,7 @@ retry: err = drm_gpusvm_range_get_pages(&vm->svm.gpusvm, r, &ctx); /* Corner where CPU mappings have changed */ if (err == -EOPNOTSUPP || err == -EFAULT || err == -EPERM) { + ctx.timeslice_ms <<= 1; /* Double timeslice if we have to retry */ if (migrate_try_count > 0 || !ctx.devmem_only) { if (err == -EOPNOTSUPP) { range_debug(range, "PAGE FAULT - EVICT PAGES"); @@ -877,6 +881,7 @@ retry_bind: drm_exec_fini(&exec); err = PTR_ERR(fence); if (err == -EAGAIN) { + ctx.timeslice_ms <<= 1; /* Double timeslice if we have to retry */ range_debug(range, "PAGE FAULT - RETRY BIND"); goto retry; } From 66c8f7b435bddb7d8577ac8a57e175a6cb147227 Mon Sep 17 00:00:00 2001 From: Umesh Nerlige Ramappa Date: Fri, 9 May 2025 09:12:01 -0700 Subject: [PATCH 22/24] drm/xe: Save CTX_TIMESTAMP mmio value instead of LRC value For determining actual job execution time, save the current value of the CTX_TIMESTAMP register rather than the value saved in LRC since the current register value is the closest to the start time of the job. v2: Define MI_STORE_REGISTER_MEM to fix compile error v3: Place MI_STORE_REGISTER_MEM sorted by MI_INSTR (Lucas) Fixes: 65921374c48f ("drm/xe: Emit ctx timestamp copy in ring ops") Signed-off-by: Umesh Nerlige Ramappa Reviewed-by: Matthew Brost Reviewed-by: Lucas De Marchi Link: https://lore.kernel.org/r/20250509161159.2173069-6-umesh.nerlige.ramappa@intel.com (cherry picked from commit 38b14233e5deff51db8faec287b4acd227152246) Signed-off-by: Lucas De Marchi --- drivers/gpu/drm/xe/instructions/xe_mi_commands.h | 4 ++++ drivers/gpu/drm/xe/xe_lrc.c | 2 +- drivers/gpu/drm/xe/xe_ring_ops.c | 7 ++----- 3 files changed, 7 insertions(+), 6 deletions(-) diff --git a/drivers/gpu/drm/xe/instructions/xe_mi_commands.h b/drivers/gpu/drm/xe/instructions/xe_mi_commands.h index 167fb0f742de..5a47991b4b81 100644 --- a/drivers/gpu/drm/xe/instructions/xe_mi_commands.h +++ b/drivers/gpu/drm/xe/instructions/xe_mi_commands.h @@ -47,6 +47,10 @@ #define MI_LRI_FORCE_POSTED REG_BIT(12) #define MI_LRI_LEN(x) (((x) & 0xff) + 1) +#define MI_STORE_REGISTER_MEM (__MI_INSTR(0x24) | XE_INSTR_NUM_DW(4)) +#define MI_SRM_USE_GGTT REG_BIT(22) +#define MI_SRM_ADD_CS_OFFSET REG_BIT(19) + #define MI_FLUSH_DW __MI_INSTR(0x26) #define MI_FLUSH_DW_PROTECTED_MEM_EN REG_BIT(22) #define MI_FLUSH_DW_STORE_INDEX REG_BIT(21) diff --git a/drivers/gpu/drm/xe/xe_lrc.c b/drivers/gpu/drm/xe/xe_lrc.c index df3ceddede07..48e8fd85eabe 100644 --- a/drivers/gpu/drm/xe/xe_lrc.c +++ b/drivers/gpu/drm/xe/xe_lrc.c @@ -684,7 +684,7 @@ static inline u32 __xe_lrc_start_seqno_offset(struct xe_lrc *lrc) static u32 __xe_lrc_ctx_job_timestamp_offset(struct xe_lrc *lrc) { - /* The start seqno is stored in the driver-defined portion of PPHWSP */ + /* This is stored in the driver-defined portion of PPHWSP */ return xe_lrc_pphwsp_offset(lrc) + LRC_CTX_JOB_TIMESTAMP_OFFSET; } diff --git a/drivers/gpu/drm/xe/xe_ring_ops.c b/drivers/gpu/drm/xe/xe_ring_ops.c index a7582b097ae6..bc1689db4cd7 100644 --- a/drivers/gpu/drm/xe/xe_ring_ops.c +++ b/drivers/gpu/drm/xe/xe_ring_ops.c @@ -234,13 +234,10 @@ static u32 get_ppgtt_flag(struct xe_sched_job *job) static int emit_copy_timestamp(struct xe_lrc *lrc, u32 *dw, int i) { - dw[i++] = MI_COPY_MEM_MEM | MI_COPY_MEM_MEM_SRC_GGTT | - MI_COPY_MEM_MEM_DST_GGTT; + dw[i++] = MI_STORE_REGISTER_MEM | MI_SRM_USE_GGTT | MI_SRM_ADD_CS_OFFSET; + dw[i++] = RING_CTX_TIMESTAMP(0).addr; dw[i++] = xe_lrc_ctx_job_timestamp_ggtt_addr(lrc); dw[i++] = 0; - dw[i++] = xe_lrc_ctx_timestamp_ggtt_addr(lrc); - dw[i++] = 0; - dw[i++] = MI_NOOP; return i; } From ce15563e49fb0b5c802564433ff8468acd1339eb Mon Sep 17 00:00:00 2001 From: Umesh Nerlige Ramappa Date: Fri, 9 May 2025 09:12:02 -0700 Subject: [PATCH 23/24] drm/xe: Save the gt pointer in lrc and drop the tile Save the gt pointer in the lrc so that it can used for gt based helpers. Signed-off-by: Umesh Nerlige Ramappa Reviewed-by: Matthew Brost Reviewed-by: Lucas De Marchi Link: https://lore.kernel.org/r/20250509161159.2173069-7-umesh.nerlige.ramappa@intel.com (cherry picked from commit 741d3ef8b8b88fab2729ca89de1180e49bc9cef0) Signed-off-by: Lucas De Marchi --- drivers/gpu/drm/xe/xe_lrc.c | 4 ++-- drivers/gpu/drm/xe/xe_lrc_types.h | 4 ++-- 2 files changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/gpu/drm/xe/xe_lrc.c b/drivers/gpu/drm/xe/xe_lrc.c index 48e8fd85eabe..79b43fec4f79 100644 --- a/drivers/gpu/drm/xe/xe_lrc.c +++ b/drivers/gpu/drm/xe/xe_lrc.c @@ -864,7 +864,7 @@ static void *empty_lrc_data(struct xe_hw_engine *hwe) static void xe_lrc_set_ppgtt(struct xe_lrc *lrc, struct xe_vm *vm) { - u64 desc = xe_vm_pdp4_descriptor(vm, lrc->tile); + u64 desc = xe_vm_pdp4_descriptor(vm, gt_to_tile(lrc->gt)); xe_lrc_write_ctx_reg(lrc, CTX_PDP0_UDW, upper_32_bits(desc)); xe_lrc_write_ctx_reg(lrc, CTX_PDP0_LDW, lower_32_bits(desc)); @@ -896,6 +896,7 @@ static int xe_lrc_init(struct xe_lrc *lrc, struct xe_hw_engine *hwe, int err; kref_init(&lrc->refcount); + lrc->gt = gt; lrc->flags = 0; lrc_size = ring_size + xe_gt_lrc_size(gt, hwe->class); if (xe_gt_has_indirect_ring_state(gt)) @@ -914,7 +915,6 @@ static int xe_lrc_init(struct xe_lrc *lrc, struct xe_hw_engine *hwe, return PTR_ERR(lrc->bo); lrc->size = lrc_size; - lrc->tile = gt_to_tile(hwe->gt); lrc->ring.size = ring_size; lrc->ring.tail = 0; lrc->ctx_timestamp = 0; diff --git a/drivers/gpu/drm/xe/xe_lrc_types.h b/drivers/gpu/drm/xe/xe_lrc_types.h index 71ecb453f811..cd38586ae989 100644 --- a/drivers/gpu/drm/xe/xe_lrc_types.h +++ b/drivers/gpu/drm/xe/xe_lrc_types.h @@ -25,8 +25,8 @@ struct xe_lrc { /** @size: size of lrc including any indirect ring state page */ u32 size; - /** @tile: tile which this LRC belongs to */ - struct xe_tile *tile; + /** @gt: gt which this LRC belongs to */ + struct xe_gt *gt; /** @flags: LRC flags */ #define XE_LRC_FLAG_INDIRECT_RING_STATE 0x1 From 617d824c5323b8474b3665ae6c410c98b839e0b0 Mon Sep 17 00:00:00 2001 From: Umesh Nerlige Ramappa Date: Fri, 9 May 2025 09:12:03 -0700 Subject: [PATCH 24/24] drm/xe: Add WA BB to capture active context utilization Context Timestamp (CTX_TIMESTAMP) in the LRC accumulates the run ticks of the context, but only gets updated when the context switches out. In order to check how long a context has been active before it switches out, two things are required: (1) Determine if the context is running: To do so, we program the WA BB to set an initial value for CTX_TIMESTAMP in the LRC. The value chosen is 1 since 0 is the initial value when the LRC is initialized. During a query, we just check for this value to determine if the context is active. If the context switched out, it would overwrite this location with the actual CTX_TIMESTAMP MMIO value. Note that WA BB runs as the last part of the context restore, so reusing this LRC location will not clobber anything. (2) Calculate the time that the context has been active for: The CTX_TIMESTAMP ticks only when the context is active. If a context is active, we just use the CTX_TIMESTAMP MMIO as the new value of utilization. While doing so, we need to read the CTX_TIMESTAMP MMIO for the specific engine instance. Since we do not know which instance the context is running on until it is scheduled, we also read the ENGINE_ID MMIO in the WA BB and store it in the PPHSWP. Using the above 2 instructions in a WA BB, capture active context utilization. v2: (Matt Brost) - This breaks TDR, fix it by saving the CTX_TIMESTAMP register "drm/xe: Save CTX_TIMESTAMP mmio value instead of LRC value" - Drop tile from LRC if using gt "drm/xe: Save the gt pointer in LRC and drop the tile" v3: - Remove helpers for bb_per_ctx_ptr (Matt) - Add define for context active value (Matt) - Use 64 bit CTX TIMESTAMP for platforms that support it. For platforms that don't, live with the rare race. (Matt, Lucas) - Convert engine id to hwe and get the MMIO value (Lucas) - Correct commit message on when WA BB runs (Lucas) v4: - s/GRAPHICS_VER(...)/xe->info.has_64bit_timestamp/ (Matt) - Drop support for active utilization on a VF (CI failure) - In xe_lrc_init ensure the lrc value is 0 to begin with (CI regression) v5: - Minor checkpatch fix - Squash into previous commit and make TDR use 32-bit time - Update code comment to match commit msg Closes: https://gitlab.freedesktop.org/drm/xe/kernel/-/issues/4532 Cc: # v6.13+ Suggested-by: Lucas De Marchi Signed-off-by: Umesh Nerlige Ramappa Reviewed-by: Lucas De Marchi Link: https://lore.kernel.org/r/20250509161159.2173069-8-umesh.nerlige.ramappa@intel.com (cherry picked from commit 82b98cadb01f63cdb159e596ec06866d00f8e8c7) Signed-off-by: Lucas De Marchi --- drivers/gpu/drm/xe/regs/xe_engine_regs.h | 5 + drivers/gpu/drm/xe/regs/xe_lrc_layout.h | 2 + drivers/gpu/drm/xe/xe_device_types.h | 2 + drivers/gpu/drm/xe/xe_exec_queue.c | 2 +- drivers/gpu/drm/xe/xe_guc_submit.c | 2 +- drivers/gpu/drm/xe/xe_lrc.c | 193 +++++++++++++++++++++-- drivers/gpu/drm/xe/xe_lrc.h | 5 +- drivers/gpu/drm/xe/xe_lrc_types.h | 5 +- drivers/gpu/drm/xe/xe_pci.c | 2 + drivers/gpu/drm/xe/xe_pci_types.h | 1 + drivers/gpu/drm/xe/xe_trace_lrc.h | 8 +- 11 files changed, 208 insertions(+), 19 deletions(-) diff --git a/drivers/gpu/drm/xe/regs/xe_engine_regs.h b/drivers/gpu/drm/xe/regs/xe_engine_regs.h index fb8ec317b6ee..891f928d80ce 100644 --- a/drivers/gpu/drm/xe/regs/xe_engine_regs.h +++ b/drivers/gpu/drm/xe/regs/xe_engine_regs.h @@ -43,6 +43,10 @@ #define XEHPC_BCS8_RING_BASE 0x3ee000 #define GSCCS_RING_BASE 0x11a000 +#define ENGINE_ID(base) XE_REG((base) + 0x8c) +#define ENGINE_INSTANCE_ID REG_GENMASK(9, 4) +#define ENGINE_CLASS_ID REG_GENMASK(2, 0) + #define RING_TAIL(base) XE_REG((base) + 0x30) #define TAIL_ADDR REG_GENMASK(20, 3) @@ -154,6 +158,7 @@ #define STOP_RING REG_BIT(8) #define RING_CTX_TIMESTAMP(base) XE_REG((base) + 0x3a8) +#define RING_CTX_TIMESTAMP_UDW(base) XE_REG((base) + 0x3ac) #define CSBE_DEBUG_STATUS(base) XE_REG((base) + 0x3fc) #define RING_FORCE_TO_NONPRIV(base, i) XE_REG(((base) + 0x4d0) + (i) * 4) diff --git a/drivers/gpu/drm/xe/regs/xe_lrc_layout.h b/drivers/gpu/drm/xe/regs/xe_lrc_layout.h index 57944f90bbf6..994af591a2e8 100644 --- a/drivers/gpu/drm/xe/regs/xe_lrc_layout.h +++ b/drivers/gpu/drm/xe/regs/xe_lrc_layout.h @@ -11,7 +11,9 @@ #define CTX_RING_TAIL (0x06 + 1) #define CTX_RING_START (0x08 + 1) #define CTX_RING_CTL (0x0a + 1) +#define CTX_BB_PER_CTX_PTR (0x12 + 1) #define CTX_TIMESTAMP (0x22 + 1) +#define CTX_TIMESTAMP_UDW (0x24 + 1) #define CTX_INDIRECT_RING_STATE (0x26 + 1) #define CTX_PDP0_UDW (0x30 + 1) #define CTX_PDP0_LDW (0x32 + 1) diff --git a/drivers/gpu/drm/xe/xe_device_types.h b/drivers/gpu/drm/xe/xe_device_types.h index 9f8667ebba85..0482f26aa480 100644 --- a/drivers/gpu/drm/xe/xe_device_types.h +++ b/drivers/gpu/drm/xe/xe_device_types.h @@ -330,6 +330,8 @@ struct xe_device { u8 has_sriov:1; /** @info.has_usm: Device has unified shared memory support */ u8 has_usm:1; + /** @info.has_64bit_timestamp: Device supports 64-bit timestamps */ + u8 has_64bit_timestamp:1; /** @info.is_dgfx: is discrete device */ u8 is_dgfx:1; /** diff --git a/drivers/gpu/drm/xe/xe_exec_queue.c b/drivers/gpu/drm/xe/xe_exec_queue.c index 606922d9dd73..cd9b1c32f30f 100644 --- a/drivers/gpu/drm/xe/xe_exec_queue.c +++ b/drivers/gpu/drm/xe/xe_exec_queue.c @@ -830,7 +830,7 @@ void xe_exec_queue_update_run_ticks(struct xe_exec_queue *q) { struct xe_device *xe = gt_to_xe(q->gt); struct xe_lrc *lrc; - u32 old_ts, new_ts; + u64 old_ts, new_ts; int idx; /* diff --git a/drivers/gpu/drm/xe/xe_guc_submit.c b/drivers/gpu/drm/xe/xe_guc_submit.c index 31bc2022bfc2..769781d577df 100644 --- a/drivers/gpu/drm/xe/xe_guc_submit.c +++ b/drivers/gpu/drm/xe/xe_guc_submit.c @@ -941,7 +941,7 @@ static bool check_timeout(struct xe_exec_queue *q, struct xe_sched_job *job) return xe_sched_invalidate_job(job, 2); } - ctx_timestamp = xe_lrc_ctx_timestamp(q->lrc[0]); + ctx_timestamp = lower_32_bits(xe_lrc_ctx_timestamp(q->lrc[0])); ctx_job_timestamp = xe_lrc_ctx_job_timestamp(q->lrc[0]); /* diff --git a/drivers/gpu/drm/xe/xe_lrc.c b/drivers/gpu/drm/xe/xe_lrc.c index 79b43fec4f79..03bfba696b37 100644 --- a/drivers/gpu/drm/xe/xe_lrc.c +++ b/drivers/gpu/drm/xe/xe_lrc.c @@ -24,6 +24,7 @@ #include "xe_hw_fence.h" #include "xe_map.h" #include "xe_memirq.h" +#include "xe_mmio.h" #include "xe_sriov.h" #include "xe_trace_lrc.h" #include "xe_vm.h" @@ -650,6 +651,7 @@ u32 xe_lrc_pphwsp_offset(struct xe_lrc *lrc) #define LRC_START_SEQNO_PPHWSP_OFFSET (LRC_SEQNO_PPHWSP_OFFSET + 8) #define LRC_CTX_JOB_TIMESTAMP_OFFSET (LRC_START_SEQNO_PPHWSP_OFFSET + 8) #define LRC_PARALLEL_PPHWSP_OFFSET 2048 +#define LRC_ENGINE_ID_PPHWSP_OFFSET 2096 #define LRC_PPHWSP_SIZE SZ_4K u32 xe_lrc_regs_offset(struct xe_lrc *lrc) @@ -694,11 +696,21 @@ static inline u32 __xe_lrc_parallel_offset(struct xe_lrc *lrc) return xe_lrc_pphwsp_offset(lrc) + LRC_PARALLEL_PPHWSP_OFFSET; } +static inline u32 __xe_lrc_engine_id_offset(struct xe_lrc *lrc) +{ + return xe_lrc_pphwsp_offset(lrc) + LRC_ENGINE_ID_PPHWSP_OFFSET; +} + static u32 __xe_lrc_ctx_timestamp_offset(struct xe_lrc *lrc) { return __xe_lrc_regs_offset(lrc) + CTX_TIMESTAMP * sizeof(u32); } +static u32 __xe_lrc_ctx_timestamp_udw_offset(struct xe_lrc *lrc) +{ + return __xe_lrc_regs_offset(lrc) + CTX_TIMESTAMP_UDW * sizeof(u32); +} + static inline u32 __xe_lrc_indirect_ring_offset(struct xe_lrc *lrc) { /* Indirect ring state page is at the very end of LRC */ @@ -726,8 +738,10 @@ DECL_MAP_ADDR_HELPERS(regs) DECL_MAP_ADDR_HELPERS(start_seqno) DECL_MAP_ADDR_HELPERS(ctx_job_timestamp) DECL_MAP_ADDR_HELPERS(ctx_timestamp) +DECL_MAP_ADDR_HELPERS(ctx_timestamp_udw) DECL_MAP_ADDR_HELPERS(parallel) DECL_MAP_ADDR_HELPERS(indirect_ring) +DECL_MAP_ADDR_HELPERS(engine_id) #undef DECL_MAP_ADDR_HELPERS @@ -742,19 +756,38 @@ u32 xe_lrc_ctx_timestamp_ggtt_addr(struct xe_lrc *lrc) return __xe_lrc_ctx_timestamp_ggtt_addr(lrc); } +/** + * xe_lrc_ctx_timestamp_udw_ggtt_addr() - Get ctx timestamp udw GGTT address + * @lrc: Pointer to the lrc. + * + * Returns: ctx timestamp udw GGTT address + */ +u32 xe_lrc_ctx_timestamp_udw_ggtt_addr(struct xe_lrc *lrc) +{ + return __xe_lrc_ctx_timestamp_udw_ggtt_addr(lrc); +} + /** * xe_lrc_ctx_timestamp() - Read ctx timestamp value * @lrc: Pointer to the lrc. * * Returns: ctx timestamp value */ -u32 xe_lrc_ctx_timestamp(struct xe_lrc *lrc) +u64 xe_lrc_ctx_timestamp(struct xe_lrc *lrc) { struct xe_device *xe = lrc_to_xe(lrc); struct iosys_map map; + u32 ldw, udw = 0; map = __xe_lrc_ctx_timestamp_map(lrc); - return xe_map_read32(xe, &map); + ldw = xe_map_read32(xe, &map); + + if (xe->info.has_64bit_timestamp) { + map = __xe_lrc_ctx_timestamp_udw_map(lrc); + udw = xe_map_read32(xe, &map); + } + + return (u64)udw << 32 | ldw; } /** @@ -877,6 +910,65 @@ static void xe_lrc_finish(struct xe_lrc *lrc) xe_bo_unpin(lrc->bo); xe_bo_unlock(lrc->bo); xe_bo_put(lrc->bo); + xe_bo_unpin_map_no_vm(lrc->bb_per_ctx_bo); +} + +/* + * xe_lrc_setup_utilization() - Setup wa bb to assist in calculating active + * context run ticks. + * @lrc: Pointer to the lrc. + * + * Context Timestamp (CTX_TIMESTAMP) in the LRC accumulates the run ticks of the + * context, but only gets updated when the context switches out. In order to + * check how long a context has been active before it switches out, two things + * are required: + * + * (1) Determine if the context is running: + * To do so, we program the WA BB to set an initial value for CTX_TIMESTAMP in + * the LRC. The value chosen is 1 since 0 is the initial value when the LRC is + * initialized. During a query, we just check for this value to determine if the + * context is active. If the context switched out, it would overwrite this + * location with the actual CTX_TIMESTAMP MMIO value. Note that WA BB runs as + * the last part of context restore, so reusing this LRC location will not + * clobber anything. + * + * (2) Calculate the time that the context has been active for: + * The CTX_TIMESTAMP ticks only when the context is active. If a context is + * active, we just use the CTX_TIMESTAMP MMIO as the new value of utilization. + * While doing so, we need to read the CTX_TIMESTAMP MMIO for the specific + * engine instance. Since we do not know which instance the context is running + * on until it is scheduled, we also read the ENGINE_ID MMIO in the WA BB and + * store it in the PPHSWP. + */ +#define CONTEXT_ACTIVE 1ULL +static void xe_lrc_setup_utilization(struct xe_lrc *lrc) +{ + u32 *cmd; + + cmd = lrc->bb_per_ctx_bo->vmap.vaddr; + + *cmd++ = MI_STORE_REGISTER_MEM | MI_SRM_USE_GGTT | MI_SRM_ADD_CS_OFFSET; + *cmd++ = ENGINE_ID(0).addr; + *cmd++ = __xe_lrc_engine_id_ggtt_addr(lrc); + *cmd++ = 0; + + *cmd++ = MI_STORE_DATA_IMM | MI_SDI_GGTT | MI_SDI_NUM_DW(1); + *cmd++ = __xe_lrc_ctx_timestamp_ggtt_addr(lrc); + *cmd++ = 0; + *cmd++ = lower_32_bits(CONTEXT_ACTIVE); + + if (lrc_to_xe(lrc)->info.has_64bit_timestamp) { + *cmd++ = MI_STORE_DATA_IMM | MI_SDI_GGTT | MI_SDI_NUM_DW(1); + *cmd++ = __xe_lrc_ctx_timestamp_udw_ggtt_addr(lrc); + *cmd++ = 0; + *cmd++ = upper_32_bits(CONTEXT_ACTIVE); + } + + *cmd++ = MI_BATCH_BUFFER_END; + + xe_lrc_write_ctx_reg(lrc, CTX_BB_PER_CTX_PTR, + xe_bo_ggtt_addr(lrc->bb_per_ctx_bo) | 1); + } #define PVC_CTX_ASID (0x2e + 1) @@ -893,6 +985,7 @@ static int xe_lrc_init(struct xe_lrc *lrc, struct xe_hw_engine *hwe, void *init_data = NULL; u32 arb_enable; u32 lrc_size; + u32 bo_flags; int err; kref_init(&lrc->refcount); @@ -902,22 +995,30 @@ static int xe_lrc_init(struct xe_lrc *lrc, struct xe_hw_engine *hwe, if (xe_gt_has_indirect_ring_state(gt)) lrc->flags |= XE_LRC_FLAG_INDIRECT_RING_STATE; + bo_flags = XE_BO_FLAG_VRAM_IF_DGFX(tile) | XE_BO_FLAG_GGTT | + XE_BO_FLAG_GGTT_INVALIDATE; + /* * FIXME: Perma-pinning LRC as we don't yet support moving GGTT address * via VM bind calls. */ lrc->bo = xe_bo_create_pin_map(xe, tile, vm, lrc_size, ttm_bo_type_kernel, - XE_BO_FLAG_VRAM_IF_DGFX(tile) | - XE_BO_FLAG_GGTT | - XE_BO_FLAG_GGTT_INVALIDATE); + bo_flags); if (IS_ERR(lrc->bo)) return PTR_ERR(lrc->bo); + lrc->bb_per_ctx_bo = xe_bo_create_pin_map(xe, tile, NULL, SZ_4K, + ttm_bo_type_kernel, + bo_flags); + if (IS_ERR(lrc->bb_per_ctx_bo)) { + err = PTR_ERR(lrc->bb_per_ctx_bo); + goto err_lrc_finish; + } + lrc->size = lrc_size; lrc->ring.size = ring_size; lrc->ring.tail = 0; - lrc->ctx_timestamp = 0; xe_hw_fence_ctx_init(&lrc->fence_ctx, hwe->gt, hwe->fence_irq, hwe->name); @@ -990,7 +1091,10 @@ static int xe_lrc_init(struct xe_lrc *lrc, struct xe_hw_engine *hwe, xe_lrc_read_ctx_reg(lrc, CTX_CONTEXT_CONTROL) | _MASKED_BIT_ENABLE(CTX_CTRL_PXP_ENABLE)); + lrc->ctx_timestamp = 0; xe_lrc_write_ctx_reg(lrc, CTX_TIMESTAMP, 0); + if (lrc_to_xe(lrc)->info.has_64bit_timestamp) + xe_lrc_write_ctx_reg(lrc, CTX_TIMESTAMP_UDW, 0); if (xe->info.has_asid && vm) xe_lrc_write_ctx_reg(lrc, PVC_CTX_ASID, vm->usm.asid); @@ -1019,6 +1123,8 @@ static int xe_lrc_init(struct xe_lrc *lrc, struct xe_hw_engine *hwe, map = __xe_lrc_start_seqno_map(lrc); xe_map_write32(lrc_to_xe(lrc), &map, lrc->fence_ctx.next_seqno - 1); + xe_lrc_setup_utilization(lrc); + return 0; err_lrc_finish: @@ -1238,6 +1344,21 @@ struct iosys_map xe_lrc_parallel_map(struct xe_lrc *lrc) return __xe_lrc_parallel_map(lrc); } +/** + * xe_lrc_engine_id() - Read engine id value + * @lrc: Pointer to the lrc. + * + * Returns: context id value + */ +static u32 xe_lrc_engine_id(struct xe_lrc *lrc) +{ + struct xe_device *xe = lrc_to_xe(lrc); + struct iosys_map map; + + map = __xe_lrc_engine_id_map(lrc); + return xe_map_read32(xe, &map); +} + static int instr_dw(u32 cmd_header) { /* GFXPIPE "SINGLE_DW" opcodes are a single dword */ @@ -1684,7 +1805,7 @@ struct xe_lrc_snapshot *xe_lrc_snapshot_capture(struct xe_lrc *lrc) snapshot->lrc_offset = xe_lrc_pphwsp_offset(lrc); snapshot->lrc_size = lrc->bo->size - snapshot->lrc_offset; snapshot->lrc_snapshot = NULL; - snapshot->ctx_timestamp = xe_lrc_ctx_timestamp(lrc); + snapshot->ctx_timestamp = lower_32_bits(xe_lrc_ctx_timestamp(lrc)); snapshot->ctx_job_timestamp = xe_lrc_ctx_job_timestamp(lrc); return snapshot; } @@ -1784,22 +1905,74 @@ void xe_lrc_snapshot_free(struct xe_lrc_snapshot *snapshot) kfree(snapshot); } +static int get_ctx_timestamp(struct xe_lrc *lrc, u32 engine_id, u64 *reg_ctx_ts) +{ + u16 class = REG_FIELD_GET(ENGINE_CLASS_ID, engine_id); + u16 instance = REG_FIELD_GET(ENGINE_INSTANCE_ID, engine_id); + struct xe_hw_engine *hwe; + u64 val; + + hwe = xe_gt_hw_engine(lrc->gt, class, instance, false); + if (xe_gt_WARN_ONCE(lrc->gt, !hwe || xe_hw_engine_is_reserved(hwe), + "Unexpected engine class:instance %d:%d for context utilization\n", + class, instance)) + return -1; + + if (lrc_to_xe(lrc)->info.has_64bit_timestamp) + val = xe_mmio_read64_2x32(&hwe->gt->mmio, + RING_CTX_TIMESTAMP(hwe->mmio_base)); + else + val = xe_mmio_read32(&hwe->gt->mmio, + RING_CTX_TIMESTAMP(hwe->mmio_base)); + + *reg_ctx_ts = val; + + return 0; +} + /** * xe_lrc_update_timestamp() - Update ctx timestamp * @lrc: Pointer to the lrc. * @old_ts: Old timestamp value * * Populate @old_ts current saved ctx timestamp, read new ctx timestamp and - * update saved value. + * update saved value. With support for active contexts, the calculation may be + * slightly racy, so follow a read-again logic to ensure that the context is + * still active before returning the right timestamp. * * Returns: New ctx timestamp value */ -u32 xe_lrc_update_timestamp(struct xe_lrc *lrc, u32 *old_ts) +u64 xe_lrc_update_timestamp(struct xe_lrc *lrc, u64 *old_ts) { + u64 lrc_ts, reg_ts; + u32 engine_id; + *old_ts = lrc->ctx_timestamp; - lrc->ctx_timestamp = xe_lrc_ctx_timestamp(lrc); + lrc_ts = xe_lrc_ctx_timestamp(lrc); + /* CTX_TIMESTAMP mmio read is invalid on VF, so return the LRC value */ + if (IS_SRIOV_VF(lrc_to_xe(lrc))) { + lrc->ctx_timestamp = lrc_ts; + goto done; + } + if (lrc_ts == CONTEXT_ACTIVE) { + engine_id = xe_lrc_engine_id(lrc); + if (!get_ctx_timestamp(lrc, engine_id, ®_ts)) + lrc->ctx_timestamp = reg_ts; + + /* read lrc again to ensure context is still active */ + lrc_ts = xe_lrc_ctx_timestamp(lrc); + } + + /* + * If context switched out, just use the lrc_ts. Note that this needs to + * be a separate if condition. + */ + if (lrc_ts != CONTEXT_ACTIVE) + lrc->ctx_timestamp = lrc_ts; + +done: trace_xe_lrc_update_timestamp(lrc, *old_ts); return lrc->ctx_timestamp; diff --git a/drivers/gpu/drm/xe/xe_lrc.h b/drivers/gpu/drm/xe/xe_lrc.h index 0b40f349ab95..eb6e8de8c939 100644 --- a/drivers/gpu/drm/xe/xe_lrc.h +++ b/drivers/gpu/drm/xe/xe_lrc.h @@ -120,7 +120,8 @@ void xe_lrc_snapshot_print(struct xe_lrc_snapshot *snapshot, struct drm_printer void xe_lrc_snapshot_free(struct xe_lrc_snapshot *snapshot); u32 xe_lrc_ctx_timestamp_ggtt_addr(struct xe_lrc *lrc); -u32 xe_lrc_ctx_timestamp(struct xe_lrc *lrc); +u32 xe_lrc_ctx_timestamp_udw_ggtt_addr(struct xe_lrc *lrc); +u64 xe_lrc_ctx_timestamp(struct xe_lrc *lrc); u32 xe_lrc_ctx_job_timestamp_ggtt_addr(struct xe_lrc *lrc); u32 xe_lrc_ctx_job_timestamp(struct xe_lrc *lrc); @@ -136,6 +137,6 @@ u32 xe_lrc_ctx_job_timestamp(struct xe_lrc *lrc); * * Returns the current LRC timestamp */ -u32 xe_lrc_update_timestamp(struct xe_lrc *lrc, u32 *old_ts); +u64 xe_lrc_update_timestamp(struct xe_lrc *lrc, u64 *old_ts); #endif diff --git a/drivers/gpu/drm/xe/xe_lrc_types.h b/drivers/gpu/drm/xe/xe_lrc_types.h index cd38586ae989..ae24cf6f8dd9 100644 --- a/drivers/gpu/drm/xe/xe_lrc_types.h +++ b/drivers/gpu/drm/xe/xe_lrc_types.h @@ -52,7 +52,10 @@ struct xe_lrc { struct xe_hw_fence_ctx fence_ctx; /** @ctx_timestamp: readout value of CTX_TIMESTAMP on last update */ - u32 ctx_timestamp; + u64 ctx_timestamp; + + /** @bb_per_ctx_bo: buffer object for per context batch wa buffer */ + struct xe_bo *bb_per_ctx_bo; }; struct xe_lrc_snapshot; diff --git a/drivers/gpu/drm/xe/xe_pci.c b/drivers/gpu/drm/xe/xe_pci.c index 818f023166d5..f4d108dc49b1 100644 --- a/drivers/gpu/drm/xe/xe_pci.c +++ b/drivers/gpu/drm/xe/xe_pci.c @@ -140,6 +140,7 @@ static const struct xe_graphics_desc graphics_xelpg = { .has_indirect_ring_state = 1, \ .has_range_tlb_invalidation = 1, \ .has_usm = 1, \ + .has_64bit_timestamp = 1, \ .va_bits = 48, \ .vm_max_level = 4, \ .hw_engine_mask = \ @@ -668,6 +669,7 @@ static int xe_info_init(struct xe_device *xe, xe->info.has_range_tlb_invalidation = graphics_desc->has_range_tlb_invalidation; xe->info.has_usm = graphics_desc->has_usm; + xe->info.has_64bit_timestamp = graphics_desc->has_64bit_timestamp; for_each_remote_tile(tile, xe, id) { int err; diff --git a/drivers/gpu/drm/xe/xe_pci_types.h b/drivers/gpu/drm/xe/xe_pci_types.h index e9b9bbc138d3..ca6b10d35573 100644 --- a/drivers/gpu/drm/xe/xe_pci_types.h +++ b/drivers/gpu/drm/xe/xe_pci_types.h @@ -21,6 +21,7 @@ struct xe_graphics_desc { u8 has_indirect_ring_state:1; u8 has_range_tlb_invalidation:1; u8 has_usm:1; + u8 has_64bit_timestamp:1; }; struct xe_media_desc { diff --git a/drivers/gpu/drm/xe/xe_trace_lrc.h b/drivers/gpu/drm/xe/xe_trace_lrc.h index 5c669a0b2180..d525cbee1e34 100644 --- a/drivers/gpu/drm/xe/xe_trace_lrc.h +++ b/drivers/gpu/drm/xe/xe_trace_lrc.h @@ -19,12 +19,12 @@ #define __dev_name_lrc(lrc) dev_name(gt_to_xe((lrc)->fence_ctx.gt)->drm.dev) TRACE_EVENT(xe_lrc_update_timestamp, - TP_PROTO(struct xe_lrc *lrc, uint32_t old), + TP_PROTO(struct xe_lrc *lrc, uint64_t old), TP_ARGS(lrc, old), TP_STRUCT__entry( __field(struct xe_lrc *, lrc) - __field(u32, old) - __field(u32, new) + __field(u64, old) + __field(u64, new) __string(name, lrc->fence_ctx.name) __string(device_id, __dev_name_lrc(lrc)) ), @@ -36,7 +36,7 @@ TRACE_EVENT(xe_lrc_update_timestamp, __assign_str(name); __assign_str(device_id); ), - TP_printk("lrc=:%p lrc->name=%s old=%u new=%u device_id:%s", + TP_printk("lrc=:%p lrc->name=%s old=%llu new=%llu device_id:%s", __entry->lrc, __get_str(name), __entry->old, __entry->new, __get_str(device_id))