November 2023 - Linux-stable-mirror

[Regression] Linux-6.6.2: SRSO kernel log messages incorrect

by Caleb Jorden

Hello, I have noticed on my zen3 and zen4 machines (Ryzen 9 5950x & Ryzen 9 7950x) that the kernel boot log regarding SRSO is suddenly incorrect with the 6.6.2 kernel. I have observed this on a completely mainline/stable kernel that I compile regularly for my machines. With the 6.6.1 and 6.7-rc2 kernels, I see correct boot messages like this: [ 0.161327] Spectre V1 : Mitigation: usercopy/swapgs barriers and __user pointer sanitization [ 0.161329] Spectre V2 : Mitigation: Retpolines [ 0.161330] Spectre V2 : Spectre v2 / SpectreRSB mitigation: Filling RSB on context switch [ 0.161331] Spectre V2 : Spectre v2 / SpectreRSB : Filling RSB on VMEXIT [ 0.161332] Spectre V2 : Enabling Restricted Speculation for firmware calls [ 0.161333] Spectre V2 : mitigation: Enabling conditional Indirect Branch Prediction Barrier [ 0.161335] Spectre V2 : User space: Mitigation: STIBP always-on protection [ 0.161336] Speculative Store Bypass: Mitigation: Speculative Store Bypass disabled via prctl [ 0.161338] Speculative Return Stack Overflow: Mitigation: safe RET However, with the 6.6.2 kernel, I see this: [ 0.164266] Spectre V1 : Mitigation: usercopy/swapgs barriers and __user pointer sanitization [ 0.164268] Spectre V2 : Mitigation: Retpolines [ 0.164269] Spectre V2 : Spectre v2 / SpectreRSB mitigation: Filling RSB on context switch [ 0.164270] Spectre V2 : Spectre v2 / SpectreRSB : Filling RSB on VMEXIT [ 0.164272] Spectre V2 : Enabling Restricted Speculation for firmware calls [ 0.164273] Spectre V2 : mitigation: Enabling conditional Indirect Branch Prediction Barrier [ 0.164275] Spectre V2 : User space: Mitigation: STIBP always-on protection [ 0.164276] Speculative Store Bypass: Mitigation: Speculative Store Bypass disabled via prctl [ 0.164278] Speculative Return Stack Overflow: IBPB-extending microcode not applied! [ 0.164279] Speculative Return Stack Overflow: WARNING: See https://kernel.org/doc/html/latest/admin-guide/hw-vuln/srso.html for mitigation options. [ 0.164280] Speculative Return Stack Overflow: Mitigation: Safe RET Notice that in both cases the final SRSO message indicates (correctly) that my system is protected with Safe RET (because my BIOS has the updated microcode for SRSO). However, in the 6.6.2 kernel I also get the microcode warning. I tracked this down to, what I believe is, the following commit from the mainline kernel not being included in the 6.6.2 patch set: 351236947a45a512c517153bbe109fe868d05e6d x86/srso: Move retbleed IBPB check into existing 'has_microcode' code block When I cherry-pick this commit to the 6.6.2 release, the log messages are correct. Note that this patch does not obviously indicate there is a functional change introduced by applying it. However (at least in the case of Zen3 and Zen4) when the updated microcode has been applied, the flow is different (specifically the situation that falls into the else statement that produces the pr-warn calls to indicate that the microcode fix needs applied). I believe this might be because RETBLEED does not apply to Zen3 and Zen4. Unfortunately I don't have a Zen1 or Zen2 system readily available on which to test this theory. While this should only be a cosmetic change, I believe that there is value in having the correct messages in the kernel logs for SRSO in this LTS release. Please feel free to correct my analysis above if it is incorrect. Warm Regards, Caleb Jorden

1 year, 1 month

3
2
0 0

FAILED: patch "[PATCH] media: qcom: camss: Fix set CSI2_RX_CFG1_VC_MODE when VC is" failed to apply to 6.1-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 6.1-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. To reproduce the conflict and resubmit, you may use the following commands: git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-6.1.y git checkout FETCH_HEAD git cherry-pick -x e655d1ae9703286cef7fda8675cad62f649dc183 # <resolve conflicts, build, test, etc.> git commit -s git send-email --to '<stable(a)vger.kernel.org>' --in-reply-to '2023112438-resistant-countdown-f624@gregkh' --subject-prefix 'PATCH 6.1.y' HEAD^.. Possible dependencies: e655d1ae9703 ("media: qcom: camss: Fix set CSI2_RX_CFG1_VC_MODE when VC is greater than 3") thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From e655d1ae9703286cef7fda8675cad62f649dc183 Mon Sep 17 00:00:00 2001 From: Bryan O'Donoghue <bryan.odonoghue(a)linaro.org> Date: Wed, 30 Aug 2023 16:16:14 +0100 Subject: [PATCH] media: qcom: camss: Fix set CSI2_RX_CFG1_VC_MODE when VC is greater than 3 VC_MODE = 0 implies a two bit VC address. VC_MODE = 1 is required for VCs with a larger address than two bits. Fixes: eebe6d00e9bf ("media: camss: Add support for CSID hardware version Titan 170") Cc: stable(a)vger.kernel.org Signed-off-by: Bryan O'Donoghue <bryan.odonoghue(a)linaro.org> Reviewed-by: Laurent Pinchart <laurent.pinchart(a)ideasonboard.com> Signed-off-by: Hans Verkuil <hverkuil-cisco(a)xs4all.nl> diff --git a/drivers/media/platform/qcom/camss/camss-csid-gen2.c b/drivers/media/platform/qcom/camss/camss-csid-gen2.c index 0f8ac29d038d..efc68f8b4de9 100644 --- a/drivers/media/platform/qcom/camss/camss-csid-gen2.c +++ b/drivers/media/platform/qcom/camss/camss-csid-gen2.c @@ -449,6 +449,8 @@ static void __csid_configure_stream(struct csid_device *csid, u8 enable, u8 vc) writel_relaxed(val, csid->base + CSID_CSI2_RX_CFG0); val = 1 << CSI2_RX_CFG1_PACKET_ECC_CORRECTION_EN; + if (vc > 3) + val |= 1 << CSI2_RX_CFG1_VC_MODE; val |= 1 << CSI2_RX_CFG1_MISR_EN; writel_relaxed(val, csid->base + CSID_CSI2_RX_CFG1);

1 year, 1 month

1
0
0 0

FAILED: patch "[PATCH] drm/amd/display: Remove wait while locked" failed to apply to 6.5-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 6.5-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. To reproduce the conflict and resubmit, you may use the following commands: git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-6.5.y git checkout FETCH_HEAD git cherry-pick -x d12f00c91fdfe3e50747f9e7e229fd8ede16b632 # <resolve conflicts, build, test, etc.> git commit -s git send-email --to '<stable(a)vger.kernel.org>' --in-reply-to '2023112444-ribbon-phobia-e4f6@gregkh' --subject-prefix 'PATCH 6.5.y' HEAD^.. Possible dependencies: d12f00c91fdf ("drm/amd/display: Remove wait while locked") thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From d12f00c91fdfe3e50747f9e7e229fd8ede16b632 Mon Sep 17 00:00:00 2001 From: Gabe Teeger <gabe.teeger(a)amd.com> Date: Mon, 14 Aug 2023 16:06:18 -0400 Subject: [PATCH] drm/amd/display: Remove wait while locked [Why] We wait for mpc idle while in a locked state, leading to potential deadlock. [What] Move the wait_for_idle call to outside of HW lock. This and a call to wait_drr_doublebuffer_pending_clear are moved added to a new static helper function called wait_for_outstanding_hw_updates, to make the interface clearer. Cc: stable(a)vger.kernel.org Fixes: 8f0d304d21b3 ("drm/amd/display: Do not commit pipe when updating DRR") Reviewed-by: Jun Lei <jun.lei(a)amd.com> Acked-by: Hamza Mahfooz <hamza.mahfooz(a)amd.com> Signed-off-by: Gabe Teeger <gabe.teeger(a)amd.com> Signed-off-by: Alex Deucher <alexander.deucher(a)amd.com> diff --git a/drivers/gpu/drm/amd/display/dc/Makefile b/drivers/gpu/drm/amd/display/dc/Makefile index 69ffd4424dc7..1b8c2aef4633 100644 --- a/drivers/gpu/drm/amd/display/dc/Makefile +++ b/drivers/gpu/drm/amd/display/dc/Makefile @@ -78,3 +78,4 @@ DC_EDID += dc_edid_parser.o AMD_DISPLAY_DMUB = $(addprefix $(AMDDALPATH)/dc/,$(DC_DMUB)) AMD_DISPLAY_EDID = $(addprefix $(AMDDALPATH)/dc/,$(DC_EDID)) AMD_DISPLAY_FILES += $(AMD_DISPLAY_DMUB) $(AMD_DISPLAY_EDID) + diff --git a/drivers/gpu/drm/amd/display/dc/core/dc.c b/drivers/gpu/drm/amd/display/dc/core/dc.c index 025e0fdf486d..c8f301aabcf8 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc.c @@ -3501,6 +3501,45 @@ static void commit_planes_for_stream_fast(struct dc *dc, top_pipe_to_program->stream->update_flags.raw = 0; } +static void wait_for_outstanding_hw_updates(struct dc *dc, const struct dc_state *dc_context) +{ +/* + * This function calls HWSS to wait for any potentially double buffered + * operations to complete. It should be invoked as a pre-amble prior + * to full update programming before asserting any HW locks. + */ + int pipe_idx; + int opp_inst; + int opp_count = dc->res_pool->pipe_count; + struct hubp *hubp; + int mpcc_inst; + const struct pipe_ctx *pipe_ctx; + + for (pipe_idx = 0; pipe_idx < dc->res_pool->pipe_count; pipe_idx++) { + pipe_ctx = &dc_context->res_ctx.pipe_ctx[pipe_idx]; + + if (!pipe_ctx->stream) + continue; + + if (pipe_ctx->stream_res.tg->funcs->wait_drr_doublebuffer_pending_clear) + pipe_ctx->stream_res.tg->funcs->wait_drr_doublebuffer_pending_clear(pipe_ctx->stream_res.tg); + + hubp = pipe_ctx->plane_res.hubp; + if (!hubp) + continue; + + mpcc_inst = hubp->inst; + // MPCC inst is equal to pipe index in practice + for (opp_inst = 0; opp_inst < opp_count; opp_inst++) { + if (dc->res_pool->opps[opp_inst]->mpcc_disconnect_pending[mpcc_inst]) { + dc->res_pool->mpc->funcs->wait_for_idle(dc->res_pool->mpc, mpcc_inst); + dc->res_pool->opps[opp_inst]->mpcc_disconnect_pending[mpcc_inst] = false; + break; + } + } + } +} + static void commit_planes_for_stream(struct dc *dc, struct dc_surface_update *srf_updates, int surface_count, @@ -3519,24 +3558,9 @@ static void commit_planes_for_stream(struct dc *dc, // dc->current_state anymore, so we have to cache it before we apply // the new SubVP context subvp_prev_use = false; - - dc_z10_restore(dc); - - if (update_type == UPDATE_TYPE_FULL) { - /* wait for all double-buffer activity to clear on all pipes */ - int pipe_idx; - - for (pipe_idx = 0; pipe_idx < dc->res_pool->pipe_count; pipe_idx++) { - struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[pipe_idx]; - - if (!pipe_ctx->stream) - continue; - - if (pipe_ctx->stream_res.tg->funcs->wait_drr_doublebuffer_pending_clear) - pipe_ctx->stream_res.tg->funcs->wait_drr_doublebuffer_pending_clear(pipe_ctx->stream_res.tg); - } - } + if (update_type == UPDATE_TYPE_FULL) + wait_for_outstanding_hw_updates(dc, context); if (update_type == UPDATE_TYPE_FULL) { dc_allow_idle_optimizations(dc, false); diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c index dd4c7a7faf28..971fa8bf6d1f 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c @@ -1563,17 +1563,6 @@ static void dcn20_update_dchubp_dpp( || plane_state->update_flags.bits.global_alpha_change || plane_state->update_flags.bits.per_pixel_alpha_change) { // MPCC inst is equal to pipe index in practice - int mpcc_inst = hubp->inst; - int opp_inst; - int opp_count = dc->res_pool->pipe_count; - - for (opp_inst = 0; opp_inst < opp_count; opp_inst++) { - if (dc->res_pool->opps[opp_inst]->mpcc_disconnect_pending[mpcc_inst]) { - dc->res_pool->mpc->funcs->wait_for_idle(dc->res_pool->mpc, mpcc_inst); - dc->res_pool->opps[opp_inst]->mpcc_disconnect_pending[mpcc_inst] = false; - break; - } - } hws->funcs.update_mpcc(dc, pipe_ctx); }

1 year, 1 month

1
0
0 0

FAILED: patch "[PATCH] drm/amd/display: Remove wait while locked" failed to apply to 6.1-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 6.1-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. To reproduce the conflict and resubmit, you may use the following commands: git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-6.1.y git checkout FETCH_HEAD git cherry-pick -x d12f00c91fdfe3e50747f9e7e229fd8ede16b632 # <resolve conflicts, build, test, etc.> git commit -s git send-email --to '<stable(a)vger.kernel.org>' --in-reply-to '2023112445-april-bazooka-9bce@gregkh' --subject-prefix 'PATCH 6.1.y' HEAD^.. Possible dependencies: d12f00c91fdf ("drm/amd/display: Remove wait while locked") 0baae6246307 ("drm/amd/display: Refactor fast update to use new HWSS build sequence") d205a800a66e ("drm/amd/display: Add visual confirm color support for MCLK switch") 6ba5a269cdc9 ("drm/amd/display: Update vactive margin and max vblank for fpo + vactive") 62cc621604a4 ("drm/amd/display: Check Vactive for VRR active for FPO + Vactive") 124155c0bd4a ("drm/amd/display: Add w/a to disable DP dual mode on certain ports") 87f0c16e0eeb ("drm/amd/display: Enable SubVP for high refresh rate displays") b058e3999021 ("drm/amd/display: Enable SubVP on PSR panels if single stream") ec341e0f4a02 ("drm/amd/display: add extra dc odm debug options") e3416e872f84 ("drm/amd/display: Add FAMS validation before trying to use it") e0a77e09c707 ("drm/amd/display: Add missing WA and MCLK validation") d170e938f01f ("drm/amd/display: On clock init, maintain DISPCLK freq") 0289e0ed1b9a ("drm/amd/display: Add FPO + VActive support") 0cdf91bf67b7 ("drm/amd/display: Enable FPO optimization") 53c8ed46e816 ("drm/amd/display: Conditionally enable 6.75 GBps link rate") 4ed793083afc ("drm/amd/display: Use per pipe P-State force for FPO") e8e5cc645b2d ("drm/amd/display: Add infrastructure for enabling FAMS for DCN30") ac18b610fd95 ("drm/amd/display: Enable FPO for configs that could reduce vlevel") f3f8f16b10f8 ("drm/amd/display: enable DPG when disabling plane for phantom pipe") 7bd571b274fd ("drm/amd/display: DAL to program DISPCLK WDIVIDER if PMFW doesn't") thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From d12f00c91fdfe3e50747f9e7e229fd8ede16b632 Mon Sep 17 00:00:00 2001 From: Gabe Teeger <gabe.teeger(a)amd.com> Date: Mon, 14 Aug 2023 16:06:18 -0400 Subject: [PATCH] drm/amd/display: Remove wait while locked [Why] We wait for mpc idle while in a locked state, leading to potential deadlock. [What] Move the wait_for_idle call to outside of HW lock. This and a call to wait_drr_doublebuffer_pending_clear are moved added to a new static helper function called wait_for_outstanding_hw_updates, to make the interface clearer. Cc: stable(a)vger.kernel.org Fixes: 8f0d304d21b3 ("drm/amd/display: Do not commit pipe when updating DRR") Reviewed-by: Jun Lei <jun.lei(a)amd.com> Acked-by: Hamza Mahfooz <hamza.mahfooz(a)amd.com> Signed-off-by: Gabe Teeger <gabe.teeger(a)amd.com> Signed-off-by: Alex Deucher <alexander.deucher(a)amd.com> diff --git a/drivers/gpu/drm/amd/display/dc/Makefile b/drivers/gpu/drm/amd/display/dc/Makefile index 69ffd4424dc7..1b8c2aef4633 100644 --- a/drivers/gpu/drm/amd/display/dc/Makefile +++ b/drivers/gpu/drm/amd/display/dc/Makefile @@ -78,3 +78,4 @@ DC_EDID += dc_edid_parser.o AMD_DISPLAY_DMUB = $(addprefix $(AMDDALPATH)/dc/,$(DC_DMUB)) AMD_DISPLAY_EDID = $(addprefix $(AMDDALPATH)/dc/,$(DC_EDID)) AMD_DISPLAY_FILES += $(AMD_DISPLAY_DMUB) $(AMD_DISPLAY_EDID) + diff --git a/drivers/gpu/drm/amd/display/dc/core/dc.c b/drivers/gpu/drm/amd/display/dc/core/dc.c index 025e0fdf486d..c8f301aabcf8 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc.c @@ -3501,6 +3501,45 @@ static void commit_planes_for_stream_fast(struct dc *dc, top_pipe_to_program->stream->update_flags.raw = 0; } +static void wait_for_outstanding_hw_updates(struct dc *dc, const struct dc_state *dc_context) +{ +/* + * This function calls HWSS to wait for any potentially double buffered + * operations to complete. It should be invoked as a pre-amble prior + * to full update programming before asserting any HW locks. + */ + int pipe_idx; + int opp_inst; + int opp_count = dc->res_pool->pipe_count; + struct hubp *hubp; + int mpcc_inst; + const struct pipe_ctx *pipe_ctx; + + for (pipe_idx = 0; pipe_idx < dc->res_pool->pipe_count; pipe_idx++) { + pipe_ctx = &dc_context->res_ctx.pipe_ctx[pipe_idx]; + + if (!pipe_ctx->stream) + continue; + + if (pipe_ctx->stream_res.tg->funcs->wait_drr_doublebuffer_pending_clear) + pipe_ctx->stream_res.tg->funcs->wait_drr_doublebuffer_pending_clear(pipe_ctx->stream_res.tg); + + hubp = pipe_ctx->plane_res.hubp; + if (!hubp) + continue; + + mpcc_inst = hubp->inst; + // MPCC inst is equal to pipe index in practice + for (opp_inst = 0; opp_inst < opp_count; opp_inst++) { + if (dc->res_pool->opps[opp_inst]->mpcc_disconnect_pending[mpcc_inst]) { + dc->res_pool->mpc->funcs->wait_for_idle(dc->res_pool->mpc, mpcc_inst); + dc->res_pool->opps[opp_inst]->mpcc_disconnect_pending[mpcc_inst] = false; + break; + } + } + } +} + static void commit_planes_for_stream(struct dc *dc, struct dc_surface_update *srf_updates, int surface_count, @@ -3519,24 +3558,9 @@ static void commit_planes_for_stream(struct dc *dc, // dc->current_state anymore, so we have to cache it before we apply // the new SubVP context subvp_prev_use = false; - - dc_z10_restore(dc); - - if (update_type == UPDATE_TYPE_FULL) { - /* wait for all double-buffer activity to clear on all pipes */ - int pipe_idx; - - for (pipe_idx = 0; pipe_idx < dc->res_pool->pipe_count; pipe_idx++) { - struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[pipe_idx]; - - if (!pipe_ctx->stream) - continue; - - if (pipe_ctx->stream_res.tg->funcs->wait_drr_doublebuffer_pending_clear) - pipe_ctx->stream_res.tg->funcs->wait_drr_doublebuffer_pending_clear(pipe_ctx->stream_res.tg); - } - } + if (update_type == UPDATE_TYPE_FULL) + wait_for_outstanding_hw_updates(dc, context); if (update_type == UPDATE_TYPE_FULL) { dc_allow_idle_optimizations(dc, false); diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c index dd4c7a7faf28..971fa8bf6d1f 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c @@ -1563,17 +1563,6 @@ static void dcn20_update_dchubp_dpp( || plane_state->update_flags.bits.global_alpha_change || plane_state->update_flags.bits.per_pixel_alpha_change) { // MPCC inst is equal to pipe index in practice - int mpcc_inst = hubp->inst; - int opp_inst; - int opp_count = dc->res_pool->pipe_count; - - for (opp_inst = 0; opp_inst < opp_count; opp_inst++) { - if (dc->res_pool->opps[opp_inst]->mpcc_disconnect_pending[mpcc_inst]) { - dc->res_pool->mpc->funcs->wait_for_idle(dc->res_pool->mpc, mpcc_inst); - dc->res_pool->opps[opp_inst]->mpcc_disconnect_pending[mpcc_inst] = false; - break; - } - } hws->funcs.update_mpcc(dc, pipe_ctx); }

1 year, 1 month

1
0
0 0

FAILED: patch "[PATCH] drm/amd/display: Remove wait while locked" failed to apply to 6.6-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 6.6-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. To reproduce the conflict and resubmit, you may use the following commands: git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-6.6.y git checkout FETCH_HEAD git cherry-pick -x d12f00c91fdfe3e50747f9e7e229fd8ede16b632 # <resolve conflicts, build, test, etc.> git commit -s git send-email --to '<stable(a)vger.kernel.org>' --in-reply-to '2023112444-symphonic-earlobe-abf4@gregkh' --subject-prefix 'PATCH 6.6.y' HEAD^.. Possible dependencies: d12f00c91fdf ("drm/amd/display: Remove wait while locked") thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From d12f00c91fdfe3e50747f9e7e229fd8ede16b632 Mon Sep 17 00:00:00 2001 From: Gabe Teeger <gabe.teeger(a)amd.com> Date: Mon, 14 Aug 2023 16:06:18 -0400 Subject: [PATCH] drm/amd/display: Remove wait while locked [Why] We wait for mpc idle while in a locked state, leading to potential deadlock. [What] Move the wait_for_idle call to outside of HW lock. This and a call to wait_drr_doublebuffer_pending_clear are moved added to a new static helper function called wait_for_outstanding_hw_updates, to make the interface clearer. Cc: stable(a)vger.kernel.org Fixes: 8f0d304d21b3 ("drm/amd/display: Do not commit pipe when updating DRR") Reviewed-by: Jun Lei <jun.lei(a)amd.com> Acked-by: Hamza Mahfooz <hamza.mahfooz(a)amd.com> Signed-off-by: Gabe Teeger <gabe.teeger(a)amd.com> Signed-off-by: Alex Deucher <alexander.deucher(a)amd.com> diff --git a/drivers/gpu/drm/amd/display/dc/Makefile b/drivers/gpu/drm/amd/display/dc/Makefile index 69ffd4424dc7..1b8c2aef4633 100644 --- a/drivers/gpu/drm/amd/display/dc/Makefile +++ b/drivers/gpu/drm/amd/display/dc/Makefile @@ -78,3 +78,4 @@ DC_EDID += dc_edid_parser.o AMD_DISPLAY_DMUB = $(addprefix $(AMDDALPATH)/dc/,$(DC_DMUB)) AMD_DISPLAY_EDID = $(addprefix $(AMDDALPATH)/dc/,$(DC_EDID)) AMD_DISPLAY_FILES += $(AMD_DISPLAY_DMUB) $(AMD_DISPLAY_EDID) + diff --git a/drivers/gpu/drm/amd/display/dc/core/dc.c b/drivers/gpu/drm/amd/display/dc/core/dc.c index 025e0fdf486d..c8f301aabcf8 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc.c @@ -3501,6 +3501,45 @@ static void commit_planes_for_stream_fast(struct dc *dc, top_pipe_to_program->stream->update_flags.raw = 0; } +static void wait_for_outstanding_hw_updates(struct dc *dc, const struct dc_state *dc_context) +{ +/* + * This function calls HWSS to wait for any potentially double buffered + * operations to complete. It should be invoked as a pre-amble prior + * to full update programming before asserting any HW locks. + */ + int pipe_idx; + int opp_inst; + int opp_count = dc->res_pool->pipe_count; + struct hubp *hubp; + int mpcc_inst; + const struct pipe_ctx *pipe_ctx; + + for (pipe_idx = 0; pipe_idx < dc->res_pool->pipe_count; pipe_idx++) { + pipe_ctx = &dc_context->res_ctx.pipe_ctx[pipe_idx]; + + if (!pipe_ctx->stream) + continue; + + if (pipe_ctx->stream_res.tg->funcs->wait_drr_doublebuffer_pending_clear) + pipe_ctx->stream_res.tg->funcs->wait_drr_doublebuffer_pending_clear(pipe_ctx->stream_res.tg); + + hubp = pipe_ctx->plane_res.hubp; + if (!hubp) + continue; + + mpcc_inst = hubp->inst; + // MPCC inst is equal to pipe index in practice + for (opp_inst = 0; opp_inst < opp_count; opp_inst++) { + if (dc->res_pool->opps[opp_inst]->mpcc_disconnect_pending[mpcc_inst]) { + dc->res_pool->mpc->funcs->wait_for_idle(dc->res_pool->mpc, mpcc_inst); + dc->res_pool->opps[opp_inst]->mpcc_disconnect_pending[mpcc_inst] = false; + break; + } + } + } +} + static void commit_planes_for_stream(struct dc *dc, struct dc_surface_update *srf_updates, int surface_count, @@ -3519,24 +3558,9 @@ static void commit_planes_for_stream(struct dc *dc, // dc->current_state anymore, so we have to cache it before we apply // the new SubVP context subvp_prev_use = false; - - dc_z10_restore(dc); - - if (update_type == UPDATE_TYPE_FULL) { - /* wait for all double-buffer activity to clear on all pipes */ - int pipe_idx; - - for (pipe_idx = 0; pipe_idx < dc->res_pool->pipe_count; pipe_idx++) { - struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[pipe_idx]; - - if (!pipe_ctx->stream) - continue; - - if (pipe_ctx->stream_res.tg->funcs->wait_drr_doublebuffer_pending_clear) - pipe_ctx->stream_res.tg->funcs->wait_drr_doublebuffer_pending_clear(pipe_ctx->stream_res.tg); - } - } + if (update_type == UPDATE_TYPE_FULL) + wait_for_outstanding_hw_updates(dc, context); if (update_type == UPDATE_TYPE_FULL) { dc_allow_idle_optimizations(dc, false); diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c index dd4c7a7faf28..971fa8bf6d1f 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c @@ -1563,17 +1563,6 @@ static void dcn20_update_dchubp_dpp( || plane_state->update_flags.bits.global_alpha_change || plane_state->update_flags.bits.per_pixel_alpha_change) { // MPCC inst is equal to pipe index in practice - int mpcc_inst = hubp->inst; - int opp_inst; - int opp_count = dc->res_pool->pipe_count; - - for (opp_inst = 0; opp_inst < opp_count; opp_inst++) { - if (dc->res_pool->opps[opp_inst]->mpcc_disconnect_pending[mpcc_inst]) { - dc->res_pool->mpc->funcs->wait_for_idle(dc->res_pool->mpc, mpcc_inst); - dc->res_pool->opps[opp_inst]->mpcc_disconnect_pending[mpcc_inst] = false; - break; - } - } hws->funcs.update_mpcc(dc, pipe_ctx); }

1 year, 1 month

1
0
0 0

FAILED: patch "[PATCH] drm/amd/display: always switch off ODM before committing more" failed to apply to 6.1-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 6.1-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. To reproduce the conflict and resubmit, you may use the following commands: git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-6.1.y git checkout FETCH_HEAD git cherry-pick -x ea7e2edca8b2150f945ee25af142fef8438c9944 # <resolve conflicts, build, test, etc.> git commit -s git send-email --to '<stable(a)vger.kernel.org>' --in-reply-to '2023112431-smooth-crewmate-6ccd@gregkh' --subject-prefix 'PATCH 6.1.y' HEAD^.. Possible dependencies: ea7e2edca8b2 ("drm/amd/display: always switch off ODM before committing more streams") 24e461e84f1c ("drm/amd/display: add ODM case when looking for first split pipe") e4c1b01bc35b ("drm/amd/display: Use min transition for all SubVP plane add/remove") 9e7d03e8b046 ("drm/amd/display: Use min transition for SubVP into MPO") 1e8fd864afdc ("drm/amd/display: skip commit minimal transition state") f6ae69f49fcf ("drm/amd/display: Include surface of unaffected streams") 0e986cea0347 ("drm/amd/display: Copy DC context in the commit streams") 7b36f4d18e3e ("drm/amd/display: Enable new commit sequence only for DCN32x") 10fdb0a11c55 ("drm/amd/display: Rework context change check") 03ce7b387e8b ("drm/amd/display: Check if link state is valid") thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From ea7e2edca8b2150f945ee25af142fef8438c9944 Mon Sep 17 00:00:00 2001 From: Wenjing Liu <wenjing.liu(a)amd.com> Date: Tue, 15 Aug 2023 10:47:52 -0400 Subject: [PATCH] drm/amd/display: always switch off ODM before committing more streams ODM power optimization is only supported with single stream. When ODM power optimization is enabled, we might not have enough free pipes for enabling other stream. So when we are committing more than 1 stream we should first switch off ODM power optimization to make room for new stream and then allocating pipe resource for the new stream. Cc: stable(a)vger.kernel.org Fixes: 59de751e3845 ("drm/amd/display: add ODM case when looking for first split pipe") Reviewed-by: Dillon Varone <dillon.varone(a)amd.com> Acked-by: Hamza Mahfooz <hamza.mahfooz(a)amd.com> Signed-off-by: Wenjing Liu <wenjing.liu(a)amd.com> Signed-off-by: Alex Deucher <alexander.deucher(a)amd.com> diff --git a/drivers/gpu/drm/amd/display/dc/core/dc.c b/drivers/gpu/drm/amd/display/dc/core/dc.c index c8f301aabcf8..5aab67868cb6 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc.c @@ -2073,12 +2073,12 @@ enum dc_status dc_commit_streams(struct dc *dc, } } - /* Check for case where we are going from odm 2:1 to max - * pipe scenario. For these cases, we will call - * commit_minimal_transition_state() to exit out of odm 2:1 - * first before processing new streams + /* ODM Combine 2:1 power optimization is only applied for single stream + * scenario, it uses extra pipes than needed to reduce power consumption + * We need to switch off this feature to make room for new streams. */ - if (stream_count == dc->res_pool->pipe_count) { + if (stream_count > dc->current_state->stream_count && + dc->current_state->stream_count == 1) { for (i = 0; i < dc->res_pool->pipe_count; i++) { pipe = &dc->current_state->res_ctx.pipe_ctx[i]; if (pipe->next_odm_pipe)

1 year, 1 month

1
0
0 0

FAILED: patch "[PATCH] drm/amd/display: always switch off ODM before committing more" failed to apply to 6.5-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 6.5-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. To reproduce the conflict and resubmit, you may use the following commands: git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-6.5.y git checkout FETCH_HEAD git cherry-pick -x ea7e2edca8b2150f945ee25af142fef8438c9944 # <resolve conflicts, build, test, etc.> git commit -s git send-email --to '<stable(a)vger.kernel.org>' --in-reply-to '2023112430-curled-hurried-e134@gregkh' --subject-prefix 'PATCH 6.5.y' HEAD^.. Possible dependencies: ea7e2edca8b2 ("drm/amd/display: always switch off ODM before committing more streams") thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From ea7e2edca8b2150f945ee25af142fef8438c9944 Mon Sep 17 00:00:00 2001 From: Wenjing Liu <wenjing.liu(a)amd.com> Date: Tue, 15 Aug 2023 10:47:52 -0400 Subject: [PATCH] drm/amd/display: always switch off ODM before committing more streams ODM power optimization is only supported with single stream. When ODM power optimization is enabled, we might not have enough free pipes for enabling other stream. So when we are committing more than 1 stream we should first switch off ODM power optimization to make room for new stream and then allocating pipe resource for the new stream. Cc: stable(a)vger.kernel.org Fixes: 59de751e3845 ("drm/amd/display: add ODM case when looking for first split pipe") Reviewed-by: Dillon Varone <dillon.varone(a)amd.com> Acked-by: Hamza Mahfooz <hamza.mahfooz(a)amd.com> Signed-off-by: Wenjing Liu <wenjing.liu(a)amd.com> Signed-off-by: Alex Deucher <alexander.deucher(a)amd.com> diff --git a/drivers/gpu/drm/amd/display/dc/core/dc.c b/drivers/gpu/drm/amd/display/dc/core/dc.c index c8f301aabcf8..5aab67868cb6 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc.c @@ -2073,12 +2073,12 @@ enum dc_status dc_commit_streams(struct dc *dc, } } - /* Check for case where we are going from odm 2:1 to max - * pipe scenario. For these cases, we will call - * commit_minimal_transition_state() to exit out of odm 2:1 - * first before processing new streams + /* ODM Combine 2:1 power optimization is only applied for single stream + * scenario, it uses extra pipes than needed to reduce power consumption + * We need to switch off this feature to make room for new streams. */ - if (stream_count == dc->res_pool->pipe_count) { + if (stream_count > dc->current_state->stream_count && + dc->current_state->stream_count == 1) { for (i = 0; i < dc->res_pool->pipe_count; i++) { pipe = &dc->current_state->res_ctx.pipe_ctx[i]; if (pipe->next_odm_pipe)

1 year, 1 month

1
0
0 0

FAILED: patch "[PATCH] drm/amd/display: always switch off ODM before committing more" failed to apply to 6.6-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 6.6-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. To reproduce the conflict and resubmit, you may use the following commands: git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-6.6.y git checkout FETCH_HEAD git cherry-pick -x ea7e2edca8b2150f945ee25af142fef8438c9944 # <resolve conflicts, build, test, etc.> git commit -s git send-email --to '<stable(a)vger.kernel.org>' --in-reply-to '2023112429-sizable-fool-d6e7@gregkh' --subject-prefix 'PATCH 6.6.y' HEAD^.. Possible dependencies: ea7e2edca8b2 ("drm/amd/display: always switch off ODM before committing more streams") thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From ea7e2edca8b2150f945ee25af142fef8438c9944 Mon Sep 17 00:00:00 2001 From: Wenjing Liu <wenjing.liu(a)amd.com> Date: Tue, 15 Aug 2023 10:47:52 -0400 Subject: [PATCH] drm/amd/display: always switch off ODM before committing more streams ODM power optimization is only supported with single stream. When ODM power optimization is enabled, we might not have enough free pipes for enabling other stream. So when we are committing more than 1 stream we should first switch off ODM power optimization to make room for new stream and then allocating pipe resource for the new stream. Cc: stable(a)vger.kernel.org Fixes: 59de751e3845 ("drm/amd/display: add ODM case when looking for first split pipe") Reviewed-by: Dillon Varone <dillon.varone(a)amd.com> Acked-by: Hamza Mahfooz <hamza.mahfooz(a)amd.com> Signed-off-by: Wenjing Liu <wenjing.liu(a)amd.com> Signed-off-by: Alex Deucher <alexander.deucher(a)amd.com> diff --git a/drivers/gpu/drm/amd/display/dc/core/dc.c b/drivers/gpu/drm/amd/display/dc/core/dc.c index c8f301aabcf8..5aab67868cb6 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc.c @@ -2073,12 +2073,12 @@ enum dc_status dc_commit_streams(struct dc *dc, } } - /* Check for case where we are going from odm 2:1 to max - * pipe scenario. For these cases, we will call - * commit_minimal_transition_state() to exit out of odm 2:1 - * first before processing new streams + /* ODM Combine 2:1 power optimization is only applied for single stream + * scenario, it uses extra pipes than needed to reduce power consumption + * We need to switch off this feature to make room for new streams. */ - if (stream_count == dc->res_pool->pipe_count) { + if (stream_count > dc->current_state->stream_count && + dc->current_state->stream_count == 1) { for (i = 0; i < dc->res_pool->pipe_count; i++) { pipe = &dc->current_state->res_ctx.pipe_ctx[i]; if (pipe->next_odm_pipe)

1 year, 1 month

1
0
0 0

FAILED: patch "[PATCH] drm/amd/display: Fix 2nd DPIA encoder Assignment" failed to apply to 6.1-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 6.1-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. To reproduce the conflict and resubmit, you may use the following commands: git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-6.1.y git checkout FETCH_HEAD git cherry-pick -x 48468787c2b029e5c9f8abef777cd065647abf2e # <resolve conflicts, build, test, etc.> git commit -s git send-email --to '<stable(a)vger.kernel.org>' --in-reply-to '2023112417-author-ellipse-c521@gregkh' --subject-prefix 'PATCH 6.1.y' HEAD^.. Possible dependencies: 48468787c2b0 ("drm/amd/display: Fix 2nd DPIA encoder Assignment") 753b7e62c9cf ("drm/amd/display: Add DPIA Link Encoder Assignment Fix") 7ae1dbe6547c ("drm/amd/display: merge dc_link.h into dc.h and dc_types.h") 8e5cfe547bf3 ("drm/amd/display: upstream link_dp_dpia_bw.c") 5ca38a18b5a4 ("drm/amd/display: move public dc link function implementation to dc_link_exports") 54618888d1ea ("drm/amd/display: break down dc_link.c") 71d7e8904d54 ("drm/amd/display: Add HDMI manufacturer OUI and device id read") 65a4cfb45e0e ("drm/amdgpu/display: remove duplicate include header in files") e322843e5e33 ("drm/amd/display: fix linux dp link lost handled only one time") 0c2bfcc338eb ("drm/amd/display: Add Function declaration in dc_link") 6ca7415f11af ("drm/amd/display: merge dc_link_dp into dc_link") de3fb390175b ("drm/amd/display: move dp cts functions from dc_link_dp to link_dp_cts") c5a31f178e35 ("drm/amd/display: move dp irq handler functions from dc_link_dp to link_dp_irq_handler") e95afc1cf7c6 ("drm/amd/display: Enable AdaptiveSync in DC interface") 0078c924e733 ("drm/amd/display: move eDP panel control logic to link_edp_panel_control") bc33f5e5f05b ("drm/amd/display: create accessories, hwss and protocols sub folders in link") 2daeb74b7d66 ("drm/amdgpu/display/mst: update mst_mgr relevant variable when long HPD") 028c4ccfb812 ("drm/amd/display: force connector state when bpc changes during compliance") 603a521ec279 ("drm/amd/display: remove duplicate included header files") bd3149014dff ("drm/amd/display: Decrease messaging about DP alt mode state to debug") thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From 48468787c2b029e5c9f8abef777cd065647abf2e Mon Sep 17 00:00:00 2001 From: Mustapha Ghaddar <mghaddar(a)amd.com> Date: Tue, 22 Aug 2023 16:18:03 -0400 Subject: [PATCH] drm/amd/display: Fix 2nd DPIA encoder Assignment [HOW & Why] There seems to be an issue with 2nd DPIA acquiring link encoder for tiled displays. Solution is to remove check for eng_id before we get first dynamic encoder for it Reviewed-by: Cruise Hung <cruise.hung(a)amd.com> Reviewed-by: Meenakshikumar Somasundaram <meenakshikumar.somasundaram(a)amd.com> Cc: Mario Limonciello <mario.limonciello(a)amd.com> Cc: Alex Deucher <alexander.deucher(a)amd.com> Cc: stable(a)vger.kernel.org Acked-by: Stylon Wang <stylon.wang(a)amd.com> Signed-off-by: Mustapha Ghaddar <mghaddar(a)amd.com> Tested-by: Daniel Wheeler <daniel.wheeler(a)amd.com> Signed-off-by: Alex Deucher <alexander.deucher(a)amd.com> diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_enc_cfg.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_enc_cfg.c index b66eeac4d3d2..be5a6d008b29 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_link_enc_cfg.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_enc_cfg.c @@ -395,8 +395,7 @@ void link_enc_cfg_link_encs_assign( stream->link->dpia_preferred_eng_id != ENGINE_ID_UNKNOWN) eng_id_req = stream->link->dpia_preferred_eng_id; - if (eng_id == ENGINE_ID_UNKNOWN) - eng_id = find_first_avail_link_enc(stream->ctx, state, eng_id_req); + eng_id = find_first_avail_link_enc(stream->ctx, state, eng_id_req); } else eng_id = link_enc->preferred_engine; @@ -501,7 +500,6 @@ struct dc_link *link_enc_cfg_get_link_using_link_enc( if (stream) link = stream->link; - // dm_output_to_console("%s: No link using DIG(%d).\n", __func__, eng_id); return link; }

1 year, 1 month

1
0
0 0

FAILED: patch "[PATCH] drm/amd/display: Fix 2nd DPIA encoder Assignment" failed to apply to 6.5-stable tree

by gregkh＠linuxfoundation.org

The patch below does not apply to the 6.5-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable(a)vger.kernel.org>. To reproduce the conflict and resubmit, you may use the following commands: git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-6.5.y git checkout FETCH_HEAD git cherry-pick -x 48468787c2b029e5c9f8abef777cd065647abf2e # <resolve conflicts, build, test, etc.> git commit -s git send-email --to '<stable(a)vger.kernel.org>' --in-reply-to '2023112416-venue-tinwork-820b@gregkh' --subject-prefix 'PATCH 6.5.y' HEAD^.. Possible dependencies: 48468787c2b0 ("drm/amd/display: Fix 2nd DPIA encoder Assignment") 753b7e62c9cf ("drm/amd/display: Add DPIA Link Encoder Assignment Fix") thanks, greg k-h ------------------ original commit in Linus's tree ------------------ From 48468787c2b029e5c9f8abef777cd065647abf2e Mon Sep 17 00:00:00 2001 From: Mustapha Ghaddar <mghaddar(a)amd.com> Date: Tue, 22 Aug 2023 16:18:03 -0400 Subject: [PATCH] drm/amd/display: Fix 2nd DPIA encoder Assignment [HOW & Why] There seems to be an issue with 2nd DPIA acquiring link encoder for tiled displays. Solution is to remove check for eng_id before we get first dynamic encoder for it Reviewed-by: Cruise Hung <cruise.hung(a)amd.com> Reviewed-by: Meenakshikumar Somasundaram <meenakshikumar.somasundaram(a)amd.com> Cc: Mario Limonciello <mario.limonciello(a)amd.com> Cc: Alex Deucher <alexander.deucher(a)amd.com> Cc: stable(a)vger.kernel.org Acked-by: Stylon Wang <stylon.wang(a)amd.com> Signed-off-by: Mustapha Ghaddar <mghaddar(a)amd.com> Tested-by: Daniel Wheeler <daniel.wheeler(a)amd.com> Signed-off-by: Alex Deucher <alexander.deucher(a)amd.com> diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_enc_cfg.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_enc_cfg.c index b66eeac4d3d2..be5a6d008b29 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_link_enc_cfg.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_enc_cfg.c @@ -395,8 +395,7 @@ void link_enc_cfg_link_encs_assign( stream->link->dpia_preferred_eng_id != ENGINE_ID_UNKNOWN) eng_id_req = stream->link->dpia_preferred_eng_id; - if (eng_id == ENGINE_ID_UNKNOWN) - eng_id = find_first_avail_link_enc(stream->ctx, state, eng_id_req); + eng_id = find_first_avail_link_enc(stream->ctx, state, eng_id_req); } else eng_id = link_enc->preferred_engine; @@ -501,7 +500,6 @@ struct dc_link *link_enc_cfg_get_link_using_link_enc( if (stream) link = stream->link; - // dm_output_to_console("%s: No link using DIG(%d).\n", __func__, eng_id); return link; }

1 year, 1 month

1
0
0 0

2025

2024

2023

2022

2021

2020

2019

2018

2017

Linux-stable-mirror November 2023