From ee892ea83d99610fa33bea612de058e0955eec3a Mon Sep 17 00:00:00 2001 From: Ashutosh Dixit Date: Fri, 3 Feb 2023 07:53:09 -0800 Subject: [PATCH 1/6] drm/i915/hwmon: Enable PL1 power limit Previous documentation suggested that PL1 power limit is always enabled. However we now find this not to be the case on some platforms (such as ATSM). Therefore enable PL1 power limit during hwmon initialization. Bspec: 51864 v2: Add Bspec reference (Gwan-gyeong) v3: Add Fixes tag Fixes: 99f55efb79114 ("drm/i915/hwmon: Power PL1 limit and TDP setting") Signed-off-by: Ashutosh Dixit Reviewed-by: Gwan-gyeong Mun Signed-off-by: Rodrigo Vivi Link: https://patchwork.freedesktop.org/patch/msgid/20230203155309.1042297-1-ashutosh.dixit@intel.com (cherry picked from commit 0349c41b05968befaffa5fbb7e73d0ee6004f610) Signed-off-by: Jani Nikula --- drivers/gpu/drm/i915/i915_hwmon.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/drivers/gpu/drm/i915/i915_hwmon.c b/drivers/gpu/drm/i915/i915_hwmon.c index 1225bc432f0d..4683a5b96eff 100644 --- a/drivers/gpu/drm/i915/i915_hwmon.c +++ b/drivers/gpu/drm/i915/i915_hwmon.c @@ -687,6 +687,11 @@ hwm_get_preregistration_info(struct drm_i915_private *i915) for_each_gt(gt, i915, i) hwm_energy(&hwmon->ddat_gt[i], &energy); } + + /* Enable PL1 power limit */ + if (i915_mmio_reg_valid(hwmon->rg.pkg_rapl_limit)) + hwm_locked_with_pm_intel_uncore_rmw(ddat, hwmon->rg.pkg_rapl_limit, + PKG_PWR_LIM_1_EN, PKG_PWR_LIM_1_EN); } void i915_hwmon_register(struct drm_i915_private *i915) From 897f453c106380e57600c19a0a0485ceb4f3b0d9 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Mon, 20 Feb 2023 17:17:30 +0200 Subject: [PATCH 2/6] drm/i915: Fix audio ELD handling for DP MST MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit I forgot to call intel_audio_compute_config() on DP MST, which means ELD doesn't get populated and passed to the audio driver. References: https://gitlab.freedesktop.org/drm/intel/-/issues/8097 Fixes: 5d986635e296 ("drm/i915/audio: Precompute the ELD") Signed-off-by: Ville Syrjälä Link: https://patchwork.freedesktop.org/patch/msgid/20230220151731.6852-1-ville.syrjala@linux.intel.com Reviewed-by: Uma Shankar (cherry picked from commit 518b761a7b0e2bb2fac2518f041c71b461adf761) Signed-off-by: Jani Nikula --- drivers/gpu/drm/i915/display/intel_dp_mst.c | 25 +++++++++++++-------- 1 file changed, 16 insertions(+), 9 deletions(-) diff --git a/drivers/gpu/drm/i915/display/intel_dp_mst.c b/drivers/gpu/drm/i915/display/intel_dp_mst.c index 054a009e800d..2106b3de225a 100644 --- a/drivers/gpu/drm/i915/display/intel_dp_mst.c +++ b/drivers/gpu/drm/i915/display/intel_dp_mst.c @@ -265,6 +265,19 @@ static int intel_dp_mst_update_slots(struct intel_encoder *encoder, return 0; } +static bool intel_dp_mst_has_audio(const struct drm_connector_state *conn_state) +{ + const struct intel_digital_connector_state *intel_conn_state = + to_intel_digital_connector_state(conn_state); + struct intel_connector *connector = + to_intel_connector(conn_state->connector); + + if (intel_conn_state->force_audio == HDMI_AUDIO_AUTO) + return connector->port->has_audio; + else + return intel_conn_state->force_audio == HDMI_AUDIO_ON; +} + static int intel_dp_mst_compute_config(struct intel_encoder *encoder, struct intel_crtc_state *pipe_config, struct drm_connector_state *conn_state) @@ -272,10 +285,6 @@ static int intel_dp_mst_compute_config(struct intel_encoder *encoder, struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); struct intel_dp_mst_encoder *intel_mst = enc_to_mst(encoder); struct intel_dp *intel_dp = &intel_mst->primary->dp; - struct intel_connector *connector = - to_intel_connector(conn_state->connector); - struct intel_digital_connector_state *intel_conn_state = - to_intel_digital_connector_state(conn_state); const struct drm_display_mode *adjusted_mode = &pipe_config->hw.adjusted_mode; struct link_config_limits limits; @@ -287,11 +296,9 @@ static int intel_dp_mst_compute_config(struct intel_encoder *encoder, pipe_config->output_format = INTEL_OUTPUT_FORMAT_RGB; pipe_config->has_pch_encoder = false; - if (intel_conn_state->force_audio == HDMI_AUDIO_AUTO) - pipe_config->has_audio = connector->port->has_audio; - else - pipe_config->has_audio = - intel_conn_state->force_audio == HDMI_AUDIO_ON; + pipe_config->has_audio = + intel_dp_mst_has_audio(conn_state) && + intel_audio_compute_config(encoder, pipe_config, conn_state); /* * for MST we always configure max link bw - the spec doesn't From 71c602103c74b277bef3d20a308874a33ec8326d Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Jouni=20H=C3=B6gander?= Date: Tue, 21 Feb 2023 10:53:04 +0200 Subject: [PATCH 3/6] drm/i915/psr: Use calculated io and fast wake lines MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Currently we are using hardcoded 7 for io and fast wake lines. According to Bspec io and fast wake times are both 42us for DISPLAY_VER >= 12 and 50us and 32us for older platforms. Calculate line counts for these and configure them into PSR2_CTL accordingly Use 45 us for the fast wake calculation as 42 seems to be too tight based on testing. Bspec: 49274, 4289 Cc: Mika Kahola Cc: José Roberto de Souza Fixes: 64cf40a125ff ("drm/i915/psr: Program default IO buffer Wake and Fast Wake") Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/7725 Signed-off-by: Jouni Högander Reviewed-by: Stanislav Lisovskiy Link: https://patchwork.freedesktop.org/patch/msgid/20230221085304.3382297-1-jouni.hogander@intel.com (cherry picked from commit cb42e8ede5b475c096e473b86c356b1158b4bc3b) Signed-off-by: Jani Nikula --- .../drm/i915/display/intel_display_types.h | 2 + drivers/gpu/drm/i915/display/intel_psr.c | 78 +++++++++++++++---- 2 files changed, 63 insertions(+), 17 deletions(-) diff --git a/drivers/gpu/drm/i915/display/intel_display_types.h b/drivers/gpu/drm/i915/display/intel_display_types.h index 54c517ca9632..582234f0c49a 100644 --- a/drivers/gpu/drm/i915/display/intel_display_types.h +++ b/drivers/gpu/drm/i915/display/intel_display_types.h @@ -1631,6 +1631,8 @@ struct intel_psr { bool psr2_sel_fetch_cff_enabled; bool req_psr2_sdp_prior_scanline; u8 sink_sync_latency; + u8 io_wake_lines; + u8 fast_wake_lines; ktime_t last_entry_attempt; ktime_t last_exit; bool sink_not_reliable; diff --git a/drivers/gpu/drm/i915/display/intel_psr.c b/drivers/gpu/drm/i915/display/intel_psr.c index 7a72e15e6836..9f1a0bebae24 100644 --- a/drivers/gpu/drm/i915/display/intel_psr.c +++ b/drivers/gpu/drm/i915/display/intel_psr.c @@ -542,6 +542,14 @@ static void hsw_activate_psr2(struct intel_dp *intel_dp) val |= EDP_PSR2_FRAME_BEFORE_SU(max_t(u8, intel_dp->psr.sink_sync_latency + 1, 2)); val |= intel_psr2_get_tp_time(intel_dp); + if (DISPLAY_VER(dev_priv) >= 12) { + if (intel_dp->psr.io_wake_lines < 9 && + intel_dp->psr.fast_wake_lines < 9) + val |= TGL_EDP_PSR2_BLOCK_COUNT_NUM_2; + else + val |= TGL_EDP_PSR2_BLOCK_COUNT_NUM_3; + } + /* Wa_22012278275:adl-p */ if (IS_ADLP_DISPLAY_STEP(dev_priv, STEP_A0, STEP_E0)) { static const u8 map[] = { @@ -558,31 +566,21 @@ static void hsw_activate_psr2(struct intel_dp *intel_dp) * Still using the default IO_BUFFER_WAKE and FAST_WAKE, see * comments bellow for more information */ - u32 tmp, lines = 7; + u32 tmp; - val |= TGL_EDP_PSR2_BLOCK_COUNT_NUM_2; - - tmp = map[lines - TGL_EDP_PSR2_IO_BUFFER_WAKE_MIN_LINES]; + tmp = map[intel_dp->psr.io_wake_lines - TGL_EDP_PSR2_IO_BUFFER_WAKE_MIN_LINES]; tmp = tmp << TGL_EDP_PSR2_IO_BUFFER_WAKE_SHIFT; val |= tmp; - tmp = map[lines - TGL_EDP_PSR2_FAST_WAKE_MIN_LINES]; + tmp = map[intel_dp->psr.fast_wake_lines - TGL_EDP_PSR2_FAST_WAKE_MIN_LINES]; tmp = tmp << TGL_EDP_PSR2_FAST_WAKE_MIN_SHIFT; val |= tmp; } else if (DISPLAY_VER(dev_priv) >= 12) { - /* - * TODO: 7 lines of IO_BUFFER_WAKE and FAST_WAKE are default - * values from BSpec. In order to setting an optimal power - * consumption, lower than 4k resolution mode needs to decrease - * IO_BUFFER_WAKE and FAST_WAKE. And higher than 4K resolution - * mode needs to increase IO_BUFFER_WAKE and FAST_WAKE. - */ - val |= TGL_EDP_PSR2_BLOCK_COUNT_NUM_2; - val |= TGL_EDP_PSR2_IO_BUFFER_WAKE(7); - val |= TGL_EDP_PSR2_FAST_WAKE(7); + val |= TGL_EDP_PSR2_IO_BUFFER_WAKE(intel_dp->psr.io_wake_lines); + val |= TGL_EDP_PSR2_FAST_WAKE(intel_dp->psr.fast_wake_lines); } else if (DISPLAY_VER(dev_priv) >= 9) { - val |= EDP_PSR2_IO_BUFFER_WAKE(7); - val |= EDP_PSR2_FAST_WAKE(7); + val |= EDP_PSR2_IO_BUFFER_WAKE(intel_dp->psr.io_wake_lines); + val |= EDP_PSR2_FAST_WAKE(intel_dp->psr.fast_wake_lines); } if (intel_dp->psr.req_psr2_sdp_prior_scanline) @@ -842,6 +840,46 @@ static bool _compute_psr2_sdp_prior_scanline_indication(struct intel_dp *intel_d return true; } +static bool _compute_psr2_wake_times(struct intel_dp *intel_dp, + struct intel_crtc_state *crtc_state) +{ + struct drm_i915_private *i915 = dp_to_i915(intel_dp); + int io_wake_lines, io_wake_time, fast_wake_lines, fast_wake_time; + u8 max_wake_lines; + + if (DISPLAY_VER(i915) >= 12) { + io_wake_time = 42; + /* + * According to Bspec it's 42us, but based on testing + * it is not enough -> use 45 us. + */ + fast_wake_time = 45; + max_wake_lines = 12; + } else { + io_wake_time = 50; + fast_wake_time = 32; + max_wake_lines = 8; + } + + io_wake_lines = intel_usecs_to_scanlines( + &crtc_state->uapi.adjusted_mode, io_wake_time); + fast_wake_lines = intel_usecs_to_scanlines( + &crtc_state->uapi.adjusted_mode, fast_wake_time); + + if (io_wake_lines > max_wake_lines || + fast_wake_lines > max_wake_lines) + return false; + + if (i915->params.psr_safest_params) + io_wake_lines = fast_wake_lines = max_wake_lines; + + /* According to Bspec lower limit should be set as 7 lines. */ + intel_dp->psr.io_wake_lines = max(io_wake_lines, 7); + intel_dp->psr.fast_wake_lines = max(fast_wake_lines, 7); + + return true; +} + static bool intel_psr2_config_valid(struct intel_dp *intel_dp, struct intel_crtc_state *crtc_state) { @@ -936,6 +974,12 @@ static bool intel_psr2_config_valid(struct intel_dp *intel_dp, return false; } + if (!_compute_psr2_wake_times(intel_dp, crtc_state)) { + drm_dbg_kms(&dev_priv->drm, + "PSR2 not enabled, Unable to use long enough wake times\n"); + return false; + } + if (HAS_PSR2_SEL_FETCH(dev_priv)) { if (!intel_psr2_sel_fetch_config_valid(intel_dp, crtc_state) && !HAS_PSR_HW_TRACKING(dev_priv)) { From 46bc23dcd94569270d02c4c1f7e62ae01ebd53bb Mon Sep 17 00:00:00 2001 From: Ankit Nautiyal Date: Thu, 23 Feb 2023 10:06:19 +0530 Subject: [PATCH 4/6] drm/i915/dg2: Add HDMI pixel clock frequencies 267.30 and 319.89 MHz Add snps phy table values for HDMI pixel clocks 267.30 MHz and 319.89 MHz. Values are based on the Bspec algorithm for PLL programming for HDMI. Cc: stable@vger.kernel.org Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/8008 Signed-off-by: Ankit Nautiyal Reviewed-by: Uma Shankar Signed-off-by: Uma Shankar Link: https://patchwork.freedesktop.org/patch/msgid/20230223043619.3941382-1-ankit.k.nautiyal@intel.com (cherry picked from commit d46746b8b13cbd377ffc733e465d25800459a31b) Signed-off-by: Jani Nikula --- drivers/gpu/drm/i915/display/intel_snps_phy.c | 62 +++++++++++++++++++ 1 file changed, 62 insertions(+) diff --git a/drivers/gpu/drm/i915/display/intel_snps_phy.c b/drivers/gpu/drm/i915/display/intel_snps_phy.c index c65c771f5c46..1cfb94b5cedb 100644 --- a/drivers/gpu/drm/i915/display/intel_snps_phy.c +++ b/drivers/gpu/drm/i915/display/intel_snps_phy.c @@ -1419,6 +1419,36 @@ static const struct intel_mpllb_state dg2_hdmi_262750 = { REG_FIELD_PREP(SNPS_PHY_MPLLB_SSC_UP_SPREAD, 1), }; +static const struct intel_mpllb_state dg2_hdmi_267300 = { + .clock = 267300, + .ref_control = + REG_FIELD_PREP(SNPS_PHY_REF_CONTROL_REF_RANGE, 3), + .mpllb_cp = + REG_FIELD_PREP(SNPS_PHY_MPLLB_CP_INT, 7) | + REG_FIELD_PREP(SNPS_PHY_MPLLB_CP_PROP, 14) | + REG_FIELD_PREP(SNPS_PHY_MPLLB_CP_INT_GS, 64) | + REG_FIELD_PREP(SNPS_PHY_MPLLB_CP_PROP_GS, 124), + .mpllb_div = + REG_FIELD_PREP(SNPS_PHY_MPLLB_DIV5_CLK_EN, 1) | + REG_FIELD_PREP(SNPS_PHY_MPLLB_TX_CLK_DIV, 1) | + REG_FIELD_PREP(SNPS_PHY_MPLLB_PMIX_EN, 1) | + REG_FIELD_PREP(SNPS_PHY_MPLLB_V2I, 2) | + REG_FIELD_PREP(SNPS_PHY_MPLLB_FREQ_VCO, 3), + .mpllb_div2 = + REG_FIELD_PREP(SNPS_PHY_MPLLB_REF_CLK_DIV, 1) | + REG_FIELD_PREP(SNPS_PHY_MPLLB_MULTIPLIER, 74) | + REG_FIELD_PREP(SNPS_PHY_MPLLB_HDMI_DIV, 1), + .mpllb_fracn1 = + REG_FIELD_PREP(SNPS_PHY_MPLLB_FRACN_CGG_UPDATE_EN, 1) | + REG_FIELD_PREP(SNPS_PHY_MPLLB_FRACN_EN, 1) | + REG_FIELD_PREP(SNPS_PHY_MPLLB_FRACN_DEN, 65535), + .mpllb_fracn2 = + REG_FIELD_PREP(SNPS_PHY_MPLLB_FRACN_QUOT, 30146) | + REG_FIELD_PREP(SNPS_PHY_MPLLB_FRACN_REM, 36699), + .mpllb_sscen = + REG_FIELD_PREP(SNPS_PHY_MPLLB_SSC_UP_SPREAD, 1), +}; + static const struct intel_mpllb_state dg2_hdmi_268500 = { .clock = 268500, .ref_control = @@ -1509,6 +1539,36 @@ static const struct intel_mpllb_state dg2_hdmi_241500 = { REG_FIELD_PREP(SNPS_PHY_MPLLB_SSC_UP_SPREAD, 1), }; +static const struct intel_mpllb_state dg2_hdmi_319890 = { + .clock = 319890, + .ref_control = + REG_FIELD_PREP(SNPS_PHY_REF_CONTROL_REF_RANGE, 3), + .mpllb_cp = + REG_FIELD_PREP(SNPS_PHY_MPLLB_CP_INT, 6) | + REG_FIELD_PREP(SNPS_PHY_MPLLB_CP_PROP, 14) | + REG_FIELD_PREP(SNPS_PHY_MPLLB_CP_INT_GS, 64) | + REG_FIELD_PREP(SNPS_PHY_MPLLB_CP_PROP_GS, 124), + .mpllb_div = + REG_FIELD_PREP(SNPS_PHY_MPLLB_DIV5_CLK_EN, 1) | + REG_FIELD_PREP(SNPS_PHY_MPLLB_TX_CLK_DIV, 1) | + REG_FIELD_PREP(SNPS_PHY_MPLLB_PMIX_EN, 1) | + REG_FIELD_PREP(SNPS_PHY_MPLLB_V2I, 2) | + REG_FIELD_PREP(SNPS_PHY_MPLLB_FREQ_VCO, 2), + .mpllb_div2 = + REG_FIELD_PREP(SNPS_PHY_MPLLB_REF_CLK_DIV, 1) | + REG_FIELD_PREP(SNPS_PHY_MPLLB_MULTIPLIER, 94) | + REG_FIELD_PREP(SNPS_PHY_MPLLB_HDMI_DIV, 1), + .mpllb_fracn1 = + REG_FIELD_PREP(SNPS_PHY_MPLLB_FRACN_CGG_UPDATE_EN, 1) | + REG_FIELD_PREP(SNPS_PHY_MPLLB_FRACN_EN, 1) | + REG_FIELD_PREP(SNPS_PHY_MPLLB_FRACN_DEN, 65535), + .mpllb_fracn2 = + REG_FIELD_PREP(SNPS_PHY_MPLLB_FRACN_QUOT, 64094) | + REG_FIELD_PREP(SNPS_PHY_MPLLB_FRACN_REM, 13631), + .mpllb_sscen = + REG_FIELD_PREP(SNPS_PHY_MPLLB_SSC_UP_SPREAD, 1), +}; + static const struct intel_mpllb_state dg2_hdmi_497750 = { .clock = 497750, .ref_control = @@ -1696,8 +1756,10 @@ static const struct intel_mpllb_state * const dg2_hdmi_tables[] = { &dg2_hdmi_209800, &dg2_hdmi_241500, &dg2_hdmi_262750, + &dg2_hdmi_267300, &dg2_hdmi_268500, &dg2_hdmi_296703, + &dg2_hdmi_319890, &dg2_hdmi_497750, &dg2_hdmi_592000, &dg2_hdmi_593407, From 193c41926d152761764894f46e23b53c00186a82 Mon Sep 17 00:00:00 2001 From: Andrea Righi Date: Mon, 20 Feb 2023 18:18:58 +0100 Subject: [PATCH 5/6] drm/i915/sseu: fix max_subslices array-index-out-of-bounds access It seems that commit bc3c5e0809ae ("drm/i915/sseu: Don't try to store EU mask internally in UAPI format") exposed a potential out-of-bounds access, reported by UBSAN as following on a laptop with a gen 11 i915 card: UBSAN: array-index-out-of-bounds in drivers/gpu/drm/i915/gt/intel_sseu.c:65:27 index 6 is out of range for type 'u16 [6]' CPU: 2 PID: 165 Comm: systemd-udevd Not tainted 6.2.0-9-generic #9-Ubuntu Hardware name: Dell Inc. XPS 13 9300/077Y9N, BIOS 1.11.0 03/22/2022 Call Trace: show_stack+0x4e/0x61 dump_stack_lvl+0x4a/0x6f dump_stack+0x10/0x18 ubsan_epilogue+0x9/0x3a __ubsan_handle_out_of_bounds.cold+0x42/0x47 gen11_compute_sseu_info+0x121/0x130 [i915] intel_sseu_info_init+0x15d/0x2b0 [i915] intel_gt_init_mmio+0x23/0x40 [i915] i915_driver_mmio_probe+0x129/0x400 [i915] ? intel_gt_probe_all+0x91/0x2e0 [i915] i915_driver_probe+0xe1/0x3f0 [i915] ? drm_privacy_screen_get+0x16d/0x190 [drm] ? acpi_dev_found+0x64/0x80 i915_pci_probe+0xac/0x1b0 [i915] ... According to the definition of sseu_dev_info, eu_mask->hsw is limited to a maximum of GEN_MAX_SS_PER_HSW_SLICE (6) sub-slices, but gen11_sseu_info_init() can potentially set 8 sub-slices, in the !IS_JSL_EHL(gt->i915) case. Fix this by reserving up to 8 slots for max_subslices in the eu_mask struct. Reported-by: Emil Renner Berthing Signed-off-by: Andrea Righi Fixes: bc3c5e0809ae ("drm/i915/sseu: Don't try to store EU mask internally in UAPI format") Reviewed-by: Matt Roper Signed-off-by: Matt Roper Link: https://patchwork.freedesktop.org/patch/msgid/20230220171858.131416-1-andrea.righi@canonical.com (cherry picked from commit 3cba09a6ac86ea1d456909626eb2685596c07822) Signed-off-by: Jani Nikula --- drivers/gpu/drm/i915/gt/intel_sseu.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/gt/intel_sseu.h b/drivers/gpu/drm/i915/gt/intel_sseu.h index aa87d3832d60..d7e8c374f153 100644 --- a/drivers/gpu/drm/i915/gt/intel_sseu.h +++ b/drivers/gpu/drm/i915/gt/intel_sseu.h @@ -27,7 +27,7 @@ struct drm_printer; * is only relevant to pre-Xe_HP platforms (Xe_HP and beyond use the * I915_MAX_SS_FUSE_BITS value below). */ -#define GEN_MAX_SS_PER_HSW_SLICE 6 +#define GEN_MAX_SS_PER_HSW_SLICE 8 /* * Maximum number of 32-bit registers used by hardware to express the From e0e6b416b25ee14716f3549e0cbec1011b193809 Mon Sep 17 00:00:00 2001 From: Janusz Krzysztofik Date: Thu, 2 Mar 2023 13:08:20 +0100 Subject: [PATCH 6/6] drm/i915/active: Fix misuse of non-idle barriers as fence trackers Users reported oopses on list corruptions when using i915 perf with a number of concurrently running graphics applications. Root cause analysis pointed at an issue in barrier processing code -- a race among perf open / close replacing active barriers with perf requests on kernel context and concurrent barrier preallocate / acquire operations performed during user context first pin / last unpin. When adding a request to a composite tracker, we try to reuse an existing fence tracker, already allocated and registered with that composite. The tracker we obtain may already track another fence, may be an idle barrier, or an active barrier. If the tracker we get occurs a non-idle barrier then we try to delete that barrier from a list of barrier tasks it belongs to. However, while doing that we don't respect return value from a function that performs the barrier deletion. Should the deletion ever fail, we would end up reusing the tracker still registered as a barrier task. Since the same structure field is reused with both fence callback lists and barrier tasks list, list corruptions would likely occur. Barriers are now deleted from a barrier tasks list by temporarily removing the list content, traversing that content with skip over the node to be deleted, then populating the list back with the modified content. Should that intentionally racy concurrent deletion attempts be not serialized, one or more of those may fail because of the list being temporary empty. Related code that ignores the results of barrier deletion was initially introduced in v5.4 by commit d8af05ff38ae ("drm/i915: Allow sharing the idle-barrier from other kernel requests"). However, all users of the barrier deletion routine were apparently serialized at that time, then the issue didn't exhibit itself. Results of git bisect with help of a newly developed igt@gem_barrier_race@remote-request IGT test indicate that list corruptions might start to appear after commit 311770173fac ("drm/i915/gt: Schedule request retirement when timeline idles"), introduced in v5.5. Respect results of barrier deletion attempts -- mark the barrier as idle only if successfully deleted from the list. Then, before proceeding with setting our fence as the one currently tracked, make sure that the tracker we've got is not a non-idle barrier. If that check fails then don't use that tracker but go back and try to acquire a new, usable one. v3: use unlikely() to document what outcome we expect (Andi), - fix bad grammar in commit description. v2: no code changes, - blame commit 311770173fac ("drm/i915/gt: Schedule request retirement when timeline idles"), v5.5, not commit d8af05ff38ae ("drm/i915: Allow sharing the idle-barrier from other kernel requests"), v5.4, - reword commit description. Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/6333 Fixes: 311770173fac ("drm/i915/gt: Schedule request retirement when timeline idles") Cc: Chris Wilson Cc: stable@vger.kernel.org # v5.5 Cc: Andi Shyti Signed-off-by: Janusz Krzysztofik Reviewed-by: Andi Shyti Signed-off-by: Andi Shyti Link: https://patchwork.freedesktop.org/patch/msgid/20230302120820.48740-1-janusz.krzysztofik@linux.intel.com (cherry picked from commit 506006055769b10d1b2b4e22f636f3b45e0e9fc7) Signed-off-by: Jani Nikula --- drivers/gpu/drm/i915/i915_active.c | 25 ++++++++++++++----------- 1 file changed, 14 insertions(+), 11 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_active.c b/drivers/gpu/drm/i915/i915_active.c index 7412abf166a8..a9fea115f2d2 100644 --- a/drivers/gpu/drm/i915/i915_active.c +++ b/drivers/gpu/drm/i915/i915_active.c @@ -422,12 +422,12 @@ replace_barrier(struct i915_active *ref, struct i915_active_fence *active) * we can use it to substitute for the pending idle-barrer * request that we want to emit on the kernel_context. */ - __active_del_barrier(ref, node_from_active(active)); - return true; + return __active_del_barrier(ref, node_from_active(active)); } int i915_active_add_request(struct i915_active *ref, struct i915_request *rq) { + u64 idx = i915_request_timeline(rq)->fence_context; struct dma_fence *fence = &rq->fence; struct i915_active_fence *active; int err; @@ -437,16 +437,19 @@ int i915_active_add_request(struct i915_active *ref, struct i915_request *rq) if (err) return err; - active = active_instance(ref, i915_request_timeline(rq)->fence_context); - if (!active) { - err = -ENOMEM; - goto out; - } + do { + active = active_instance(ref, idx); + if (!active) { + err = -ENOMEM; + goto out; + } + + if (replace_barrier(ref, active)) { + RCU_INIT_POINTER(active->fence, NULL); + atomic_dec(&ref->count); + } + } while (unlikely(is_barrier(active))); - if (replace_barrier(ref, active)) { - RCU_INIT_POINTER(active->fence, NULL); - atomic_dec(&ref->count); - } if (!__i915_active_fence_set(active, fence)) __i915_active_acquire(ref);