[PATCH AUTOSEL 6.15 032/110] drm/amdgpu: Add basic validation for RAS header

[PATCH AUTOSEL 6.15 032/110] drm/amdgpu: Add basic validation for RAS header

[Sasha Levin]

From: Lijo Lazar <lijo.lazar@amd.com>

[ Upstream commit 5df0d6addb7e9b6f71f7162d1253762a5be9138e ]

If RAS header read from EEPROM is corrupted, it could result in trying
to allocate huge memory for reading the records. Add some validation to
header fields.

Signed-off-by: Lijo Lazar <lijo.lazar@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---

Based on my analysis, here is my assessment: **YES** This commit should
be backported to stable kernel trees. Here's my extensive analysis: ##
Security Vulnerability Analysis **Critical Memory Allocation
Vulnerability:** The commit fixes a serious security vulnerability where
corrupted RAS (Reliability, Availability, and Serviceability) EEPROM
header data could cause huge memory allocations. In
`/home/sasha/linux/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c:3011`, the
code performs: ```c bps = kcalloc(control->ras_num_recs, sizeof(*bps),
GFP_KERNEL); ``` **Before the fix:** If `control->ras_num_recs` contains
a corrupted large value (e.g., 0xFFFFFFFF), this would attempt to
allocate `0xFFFFFFFF 0001-Fix-Clippy-warnings.patch 0002-Enhance-
inference-prompt-to-utilize-CVEKERNELDIR-whe.patch 0003-Update-to-
latest-version-of-clap.patch Cargo.lock Cargo.toml LICENSE README.md
analyze_merge_commit.sh dpp_rcg_backport_analysis.md
drm_bridge_analysis.txt drm_imagination_register_update_analysis.md
drm_mediatek_mtk_dpi_refactoring_analysis.md io_uring_analysis.txt
ksmbd_analysis.txt merge_commit_analysis.txt model prompt src target
test_gpio_cleanup.txt test_patch.txt verisilicon_av1_4k_analysis.md 24
bytes` = ~96GB of memory, likely causing: 1. System memory exhaustion 2.
Denial of service 3. Potential system crash/instability ## Code Changes
Analysis **1. Version Validation Enhancement:** The fix replaces a
simple `if/else` with a robust `switch` statement: ```c // Before: Only
checked for version >= V2_1 if (hdr->version >= RAS_TABLE_VER_V2_1) { //
After: Explicit validation of known versions switch (hdr->version) {
case RAS_TABLE_VER_V2_1: case RAS_TABLE_VER_V3: // Future-proofing //
V2.1+ handling break; case RAS_TABLE_VER_V1: // V1 handling break;
default: dev_err(adev->dev, "RAS header invalid, unsupported version:
%u", hdr->version); return -EINVAL; } ``` **2. Record Count Bounds
Checking:** Critical addition of bounds validation: ```c if
(control->ras_num_recs > control->ras_max_record_count) {
dev_err(adev->dev, "RAS header invalid, records in header: %u max
allowed :%u", control->ras_num_recs, control->ras_max_record_count);
return -EINVAL; } ``` This prevents the memory allocation attack by
ensuring `ras_num_recs` cannot exceed reasonable bounds. ## Stable Tree
Criteria Assessment ✅ **Fixes important security bug:** Prevents DoS via
memory exhaustion ✅ **Small and contained:** Only adds validation logic,
no functional changes ✅ **Clear side effects:** None - only adds error
checking ✅ **No architectural changes:** Pure validation enhancement ✅
**Touches critical subsystem:** Graphics driver reliability/security ✅
**Minimal regression risk:** Only adds stricter validation ## Comparison
with Historical Patterns Looking at the provided similar commits (all
marked "NO"), they were architectural refactoring changes that: - Split
functions (`ras_eeprom_init into init and check functions`) - Added new
infrastructure (`Hook EEPROM table to RAS`) - Wrapped function calls
(`RAS xfer to read/write`) - Added entirely new features (`Add RAS
EEPROM table`) In contrast, this commit is a **security hardening fix**
that adds essential input validation to prevent memory exhaustion
attacks. ## Impact Assessment **Risk if NOT backported:** - Systems with
AMD GPUs vulnerable to DoS attacks via corrupted EEPROM data - Potential
for system instability when corrupted data triggers massive allocations
- Security exposure in enterprise/server environments using AMD GPUs
**Risk if backported:** - Minimal: Only adds validation, existing
correct data will continue to work - Possible rejection of previously
accepted (but actually corrupted) EEPROM data - this is desired behavior
This fix addresses a clear security vulnerability with minimal code
changes and should definitely be backported to stable trees to protect
users from memory exhaustion attacks via corrupted RAS EEPROM headers.

 .../gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c    | 22 ++++++++++++++++---
 1 file changed, 19 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c
index 0ea7cfaf3587d..e979a6086178c 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c
@@ -1392,17 +1392,33 @@ int amdgpu_ras_eeprom_init(struct amdgpu_ras_eeprom_control *control)
 
 	__decode_table_header_from_buf(hdr, buf);
 
-	if (hdr->version >= RAS_TABLE_VER_V2_1) {
+	switch (hdr->version) {
+	case RAS_TABLE_VER_V2_1:
+	case RAS_TABLE_VER_V3:
 		control->ras_num_recs = RAS_NUM_RECS_V2_1(hdr);
 		control->ras_record_offset = RAS_RECORD_START_V2_1;
 		control->ras_max_record_count = RAS_MAX_RECORD_COUNT_V2_1;
-	} else {
+		break;
+	case RAS_TABLE_VER_V1:
 		control->ras_num_recs = RAS_NUM_RECS(hdr);
 		control->ras_record_offset = RAS_RECORD_START;
 		control->ras_max_record_count = RAS_MAX_RECORD_COUNT;
+		break;
+	default:
+		dev_err(adev->dev,
+			"RAS header invalid, unsupported version: %u",
+			hdr->version);
+		return -EINVAL;
 	}
-	control->ras_fri = RAS_OFFSET_TO_INDEX(control, hdr->first_rec_offset);
 
+	if (control->ras_num_recs > control->ras_max_record_count) {
+		dev_err(adev->dev,
+			"RAS header invalid, records in header: %u max allowed :%u",
+			control->ras_num_recs, control->ras_max_record_count);
+		return -EINVAL;
+	}
+
+	control->ras_fri = RAS_OFFSET_TO_INDEX(control, hdr->first_rec_offset);
 	control->ras_num_mca_recs = 0;
 	control->ras_num_pa_recs = 0;
 	return 0;
-- 
2.39.5

[PATCH AUTOSEL 6.15 052/110] media: rkvdec: h264: Use bytesperline and buffer height as virstride

[Sasha Levin]

From: Jonas Karlman <jonas@kwiboo.se>

[ Upstream commit d5e0aa61470c48ddc04d433a00e79cef8716377a ]

Use bytesperline and buffer height to calculate the strides configured.

This does not really change anything other than ensuring the
bytesperline that is signaled to userspace matches what is configured
in HW.

Signed-off-by: Jonas Karlman <jonas@kwiboo.se>
Reviewed-by: Nicolas Dufresne <nicolas.dufresne@collabora.com>
Tested-by: Nicolas Dufresne <nicolas.dufresne@collabora.com>
Tested-by: Christopher Obbard <chris.obbard@collabora.com>
Signed-off-by: Nicolas Dufresne <nicolas.dufresne@collabora.com>
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---

### Key Reasons for NO Backport: 1. **Not a Bug Fix**: The commit
message explicitly states "This does not really change anything other
than ensuring the bytesperline that is signaled to userspace matches
what is configured in HW." This is not fixing a functional bug but
rather improving consistency between userspace and hardware
configuration. 2. **Code Quality/Consistency Improvement**: The changes
are primarily about code correctness and consistency rather than fixing
a critical issue that affects users. This aligns with similar commits
#1, #2, and #4 which were marked as NO for backporting. 3. **Staging
Driver**: The rkvdec driver is in `drivers/staging/media/`, which
indicates it's still under development and not considered stable.
Staging drivers typically have more relaxed backporting requirements. 4.
**Specific Code Changes Analysis**: - **Before**: `hor_virstride =
(sps->bit_depth_luma_minus8 + 8) 0001-Fix-Clippy-warnings.patch
0002-Enhance-inference-prompt-to-utilize-CVEKERNELDIR-whe.patch
0003-Update-to-latest-version-of-clap.patch Cargo.lock Cargo.toml
LICENSE README.md adreno_acd_support_analysis.md
amd_display_ips_sequential_ono_backport_analysis.md
analyze_merge_commit.sh dpp_rcg_backport_analysis.md
drm_amd_display_vertical_interrupt_dcn32_dcn401_backport_analysis.md
drm_bridge_analysis.txt drm_format_helper_24bit_analysis.md
drm_imagination_register_update_analysis.md
drm_mediatek_mtk_dpi_refactoring_analysis.md
intel_ipu6_constify_analysis.md io_uring_analysis.txt ksmbd_analysis.txt
merge_commit_analysis.txt model prompt src target test_gpio_cleanup.txt
test_patch.txt verisilicon_av1_4k_analysis.md dst_fmt->width / 8;` -
**After**: `hor_virstride = dst_fmt->plane_fmt[0].bytesperline;` -
**Before**: `ver_virstride = round_up(dst_fmt->height, 16);` -
**After**: `ver_virstride = dst_fmt->height;` 5. **No Risk Indication**:
The changes don't indicate they're fixing crashes, data corruption,
security issues, or other critical problems that would warrant stable
backporting. 6. **Pattern Match with Similar Commits**: Looking at the
provided examples: - Similar commits #1, #2, #4, and #5 that make
stride/calculation improvements were marked as NO - Only commit #3 that
increased max supported height (a clear functional limitation fix) was
marked as YES 7. **No Stable Tree Mention**: The commit message contains
no indication of stable tree inclusion via Fixes: tags or explicit
stable tree requests. ### Conclusion: This commit improves code
consistency by using the actual bytesperline values from userspace
instead of calculating them from SPS parameters, but it doesn't fix a
user-impacting bug. It's a code quality improvement that follows the
pattern of other NO-backport commits in the provided examples.

 drivers/staging/media/rkvdec/rkvdec-h264.c | 14 +++++++-------
 1 file changed, 7 insertions(+), 7 deletions(-)

diff --git a/drivers/staging/media/rkvdec/rkvdec-h264.c b/drivers/staging/media/rkvdec/rkvdec-h264.c
index 4fc167b42cf0c..7a1e76d423df5 100644
--- a/drivers/staging/media/rkvdec/rkvdec-h264.c
+++ b/drivers/staging/media/rkvdec/rkvdec-h264.c
@@ -896,9 +896,9 @@ static void config_registers(struct rkvdec_ctx *ctx,
 	dma_addr_t rlc_addr;
 	dma_addr_t refer_addr;
 	u32 rlc_len;
-	u32 hor_virstride = 0;
-	u32 ver_virstride = 0;
-	u32 y_virstride = 0;
+	u32 hor_virstride;
+	u32 ver_virstride;
+	u32 y_virstride;
 	u32 yuv_virstride = 0;
 	u32 offset;
 	dma_addr_t dst_addr;
@@ -909,16 +909,16 @@ static void config_registers(struct rkvdec_ctx *ctx,
 
 	f = &ctx->decoded_fmt;
 	dst_fmt = &f->fmt.pix_mp;
-	hor_virstride = (sps->bit_depth_luma_minus8 + 8) * dst_fmt->width / 8;
-	ver_virstride = round_up(dst_fmt->height, 16);
+	hor_virstride = dst_fmt->plane_fmt[0].bytesperline;
+	ver_virstride = dst_fmt->height;
 	y_virstride = hor_virstride * ver_virstride;
 
 	if (sps->chroma_format_idc == 0)
 		yuv_virstride = y_virstride;
 	else if (sps->chroma_format_idc == 1)
-		yuv_virstride += y_virstride + y_virstride / 2;
+		yuv_virstride = y_virstride + y_virstride / 2;
 	else if (sps->chroma_format_idc == 2)
-		yuv_virstride += 2 * y_virstride;
+		yuv_virstride = 2 * y_virstride;
 
 	reg = RKVDEC_Y_HOR_VIRSTRIDE(hor_virstride / 16) |
 	      RKVDEC_UV_HOR_VIRSTRIDE(hor_virstride / 16) |
-- 
2.39.5

Re: [PATCH AUTOSEL 6.15 052/110] media: rkvdec: h264: Use bytesperline and buffer height as virstride

[Greg KH]

On Sun, Jun 01, 2025 at 07:23:34PM -0400, Sasha Levin wrote:
> From: Jonas Karlman <jonas@kwiboo.se>
> 
> [ Upstream commit d5e0aa61470c48ddc04d433a00e79cef8716377a ]
> 
> Use bytesperline and buffer height to calculate the strides configured.
> 
> This does not really change anything other than ensuring the
> bytesperline that is signaled to userspace matches what is configured
> in HW.
> 
> Signed-off-by: Jonas Karlman <jonas@kwiboo.se>
> Reviewed-by: Nicolas Dufresne <nicolas.dufresne@collabora.com>
> Tested-by: Nicolas Dufresne <nicolas.dufresne@collabora.com>
> Tested-by: Christopher Obbard <chris.obbard@collabora.com>
> Signed-off-by: Nicolas Dufresne <nicolas.dufresne@collabora.com>
> Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
> Signed-off-by: Sasha Levin <sashal@kernel.org>
> ---
> 
> ### Key Reasons for NO Backport: 1. **Not a Bug Fix**: The commit
> message explicitly states "This does not really change anything other
> than ensuring the bytesperline that is signaled to userspace matches
> what is configured in HW." This is not fixing a functional bug but
> rather improving consistency between userspace and hardware
> configuration.

As the bot said "NO", why was this picked up?

thanks,

greg k-h

[PATCH AUTOSEL 6.15 053/110] media: cec: extron-da-hd-4k-plus: Fix Wformat-truncation

[Sasha Levin]

From: Hans Verkuil <hverkuil@xs4all.nl>

[ Upstream commit 5edc9b560f60c40e658af0b8e98ae2dfadc438d8 ]

Fix gcc8 warning:

drivers/media/cec/usb/extron-da-hd-4k-plus/extron-da-hd-4k-plus.c:1014:44: warning: 'DCEC' directive output may be truncated writing 4 bytes into a region of size between 0 and 53 [-Wformat-truncation=]

Resizing the 'buf' and 'cmd' arrays fixed the warning.

Signed-off-by: Ricardo Ribalda <ribalda@chromium.org>
Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---

NO This commit should not be backported to stable kernel trees for the
following reasons: 1. **Non-critical compiler warning fix**: This commit
addresses a `-Wformat-truncation` compiler warning, which is a build-
time warning and not a runtime bug that affects users. The warning
indicates potential string truncation, but looking at the code, there's
no evidence this causes actual runtime failures or security
vulnerabilities. 2. **Consistent with similar commit patterns**: Looking
at the reference commits, all the similar format-truncation warning
fixes (commits #1, #2, #3, and #5) were marked as "Backport Status: NO".
Only commit #4 was marked "YES", but that one explicitly included "Cc:
stable@vger.kernel.org" in the commit message and had a "Fixes:" tag,
which this commit lacks. 3. **Missing stable tree indicators**: The
commit message contains no "Cc: stable@vger.kernel.org" tag and no
"Fixes:" tag pointing to a specific problematic commit that introduced a
user-facing issue. 4. **Minor code quality improvement**: The changes
are purely preventive: - `char buf[CEC_MAX_MSG_SIZE 0001-Fix-Clippy-
warnings.patch 0002-Enhance-inference-prompt-to-utilize-CVEKERNELDIR-
whe.patch 0003-Update-to-latest-version-of-clap.patch Cargo.lock
Cargo.toml LICENSE README.md adreno_acd_support_analysis.md
amd_display_ips_sequential_ono_backport_analysis.md
analyze_merge_commit.sh dpp_rcg_backport_analysis.md
drm_amd_display_vertical_interrupt_dcn32_dcn401_backport_analysis.md
drm_bridge_analysis.txt drm_format_helper_24bit_analysis.md
drm_imagination_register_update_analysis.md
drm_mediatek_mtk_dpi_refactoring_analysis.md
intel_ipu6_constify_analysis.md io_uring_analysis.txt ksmbd_analysis.txt
merge_commit_analysis.txt model prompt src target test_gpio_cleanup.txt
test_patch.txt verisilicon_av1_4k_analysis.md 3 + 1];` becomes `char
buf[(CEC_MAX_MSG_SIZE - 1) 0001-Fix-Clippy-warnings.patch 0002-Enhance-
inference-prompt-to-utilize-CVEKERNELDIR-whe.patch 0003-Update-to-
latest-version-of-clap.patch Cargo.lock Cargo.toml LICENSE README.md
adreno_acd_support_analysis.md
amd_display_ips_sequential_ono_backport_analysis.md
analyze_merge_commit.sh dpp_rcg_backport_analysis.md
drm_amd_display_vertical_interrupt_dcn32_dcn401_backport_analysis.md
drm_bridge_analysis.txt drm_format_helper_24bit_analysis.md
drm_imagination_register_update_analysis.md
drm_mediatek_mtk_dpi_refactoring_analysis.md
intel_ipu6_constify_analysis.md io_uring_analysis.txt ksmbd_analysis.txt
merge_commit_analysis.txt model prompt src target test_gpio_cleanup.txt
test_patch.txt verisilicon_av1_4k_analysis.md 3 + 1];` - `char
cmd[CEC_MAX_MSG_SIZE 0001-Fix-Clippy-warnings.patch 0002-Enhance-
inference-prompt-to-utilize-CVEKERNELDIR-whe.patch 0003-Update-to-
latest-version-of-clap.patch Cargo.lock Cargo.toml LICENSE README.md
adreno_acd_support_analysis.md
amd_display_ips_sequential_ono_backport_analysis.md
analyze_merge_commit.sh dpp_rcg_backport_analysis.md
drm_amd_display_vertical_interrupt_dcn32_dcn401_backport_analysis.md
drm_bridge_analysis.txt drm_format_helper_24bit_analysis.md
drm_imagination_register_update_analysis.md
drm_mediatek_mtk_dpi_refactoring_analysis.md
intel_ipu6_constify_analysis.md io_uring_analysis.txt ksmbd_analysis.txt
merge_commit_analysis.txt model prompt src target test_gpio_cleanup.txt
test_patch.txt verisilicon_av1_4k_analysis.md 3 + 13];` becomes `char
cmd[sizeof(buf) + 14];` These changes slightly reduce buffer sizes to
better match actual usage patterns, but there's no indication the
original sizes caused buffer overflows or other runtime problems. 5.
**Media subsystem and driver-specific**: This affects a specific USB CEC
adapter driver (`extron-da-hd-4k-plus`), making it a relatively isolated
change that doesn't impact core kernel functionality or multiple
subsystems. 6. **No user-reported issues**: There's no mention of user
reports, bug fixes, or functional improvements - only compiler warning
suppression. The stable kernel tree rules prioritize important bug fixes
with minimal regression risk. This commit, while technically correct,
falls into the category of code quality improvements rather than
critical fixes that users would benefit from in stable releases.

 .../media/cec/usb/extron-da-hd-4k-plus/extron-da-hd-4k-plus.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/media/cec/usb/extron-da-hd-4k-plus/extron-da-hd-4k-plus.c b/drivers/media/cec/usb/extron-da-hd-4k-plus/extron-da-hd-4k-plus.c
index cfbfc4c1b2e67..41d019b01ec09 100644
--- a/drivers/media/cec/usb/extron-da-hd-4k-plus/extron-da-hd-4k-plus.c
+++ b/drivers/media/cec/usb/extron-da-hd-4k-plus/extron-da-hd-4k-plus.c
@@ -1002,8 +1002,8 @@ static int extron_cec_adap_transmit(struct cec_adapter *adap, u8 attempts,
 				    u32 signal_free_time, struct cec_msg *msg)
 {
 	struct extron_port *port = cec_get_drvdata(adap);
-	char buf[CEC_MAX_MSG_SIZE * 3 + 1];
-	char cmd[CEC_MAX_MSG_SIZE * 3 + 13];
+	char buf[(CEC_MAX_MSG_SIZE - 1) * 3 + 1];
+	char cmd[sizeof(buf) + 14];
 	unsigned int i;
 
 	if (port->disconnected)
-- 
2.39.5

[PATCH AUTOSEL 6.15 092/110] genirq: Retain disable depth for managed interrupts across CPU hotplug

[Sasha Levin]

From: Brian Norris <briannorris@chromium.org>

[ Upstream commit 788019eb559fd0b365f501467ceafce540e377cc ]

Affinity-managed interrupts can be shut down and restarted during CPU
hotunplug/plug. Thereby the interrupt may be left in an unexpected state.
Specifically:

 1. Interrupt is affine to CPU N
 2. disable_irq() -> depth is 1
 3. CPU N goes offline
 4. irq_shutdown() -> depth is set to 1 (again)
 5. CPU N goes online
 6. irq_startup() -> depth is set to 0 (BUG! driver expects that the interrupt
    		     	      	        still disabled)
 7. enable_irq() -> depth underflow / unbalanced enable_irq() warning

This is only a problem for managed interrupts and CPU hotplug, all other
cases like request()/free()/request() truly needs to reset a possibly stale
disable depth value.

Provide a startup function, which takes the disable depth into account, and
invoked it for the managed interrupts in the CPU hotplug path.

This requires to change irq_shutdown() to do a depth increment instead of
setting it to 1, which allows to retain the disable depth, but is harmless
for the other code paths using irq_startup(), which will still reset the
disable depth unconditionally to keep the original correct behaviour.

A kunit tests will be added separately to cover some of these aspects.

[ tglx: Massaged changelog ]

Suggested-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Brian Norris <briannorris@chromium.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/all/20250514201353.3481400-2-briannorris@chromium.org
Signed-off-by: Sasha Levin <sashal@kernel.org>
---

Based on my analysis of the commit and the current kernel code, here is
my assessment: **YES** This commit should be backported to stable kernel
trees. Here's my extensive analysis: ## Bug Analysis The commit fixes a
critical bug in the interrupt subsystem's CPU hotplug handling that
affects affinity-managed interrupts. The bug manifests as follows: 1.
**Problem Scenario**: When a managed interrupt is disabled (depth=1),
then its CPU goes offline causing `irq_shutdown()`, and later the CPU
comes back online triggering `irq_startup()`, the disable depth
accounting becomes corrupted. 2. **Current Broken Behavior** (from
existing kernel code): - `irq_shutdown()` sets `desc->depth = 1`
unconditionally (line seen in grep output) - `irq_startup()` sets
`desc->depth = 0` unconditionally - This overwrites any previous disable
state, violating the disable/enable balance 3. **User-Visible Impact**:
The bug causes: - Unbalanced `enable_irq()` warnings - Interrupts being
unexpectedly enabled during CPU hotplug - Potential system instability
or incorrect device behavior ## Code Changes Analysis The fix makes
surgical changes to preserve disable depth: ### 1. **irq_shutdown()
Change** (kernel/irq/chip.c): ```c - desc->depth = 1; + /bin /bin.usr-
is-merged /boot /dev /etc /home /init /lib /lib.usr-is-merged /lib64
/lost+found /media /mnt /opt /proc /root /run /sbin /sbin.usr-is-merged
/snap /srv /sys /tmp /usr /var +
amd_display_timing_generator_analysis.md
amd_doorbell_analysis_framework.md
amd_gpu_documentation_backport_analysis.md amd_si_chip_analysis.md
amdgpu_csb_analysis.md amdgpu_mqd_prop_backport_analysis.md
analogix_dp_backport_analysis.md cgroup_rstat_analysis.sh
cgroup_rstat_analysis_guide.md cpuset_comment_fix_backport_analysis.md
csiphy_x1e80100_backport_analysis.md
dcn315_smu_indirect_register_access_analysis.md
drm_dp_helper_dpcd_refactoring_backport_analysis.md
drm_format_helper_generic_8bit_conversion_backport_analysis.md
es8375_codec_driver_backport_analysis.md expected_findings_summary.md
hdmi_read_request_analysis.md kernel_analysis_commands.sh
kselftest_harness_teardown_metadata_backport_analysis.md linux-kernel
mediatek_vcodec_h264_backport_analysis.md
pidfd_open_kselftest_harness_compatibility_analysis.md
qualcomm_camss_bus_type_backport_analysis.md
selftests_harness_libatomic_backport_analysis.md
sphinx_version_compatibility_backport_analysis.md
spi_sh_msiof_grpmask_backport_analysis.md
spi_sh_msiof_sitmdr2_sirmdr2_bitfield_conversion_backport_analysis.md
sta2x11_removal_analysis.md test_unused_param.c test_unused_param.o
ti_bridge_encoder_crtc_refactoring_backport_analysis.md
xe_driver_flr_vf_restriction_analysis.md
xe_firmware_per_gt_backport_analysis.md
xe_sriov_logging_backport_analysis.md Increment disable depth, so that a
managed shutdown on + amd_display_timing_generator_analysis.md
amd_doorbell_analysis_framework.md
amd_gpu_documentation_backport_analysis.md amd_si_chip_analysis.md
amdgpu_csb_analysis.md amdgpu_mqd_prop_backport_analysis.md
analogix_dp_backport_analysis.md cgroup_rstat_analysis.sh
cgroup_rstat_analysis_guide.md cpuset_comment_fix_backport_analysis.md
csiphy_x1e80100_backport_analysis.md
dcn315_smu_indirect_register_access_analysis.md
drm_dp_helper_dpcd_refactoring_backport_analysis.md
drm_format_helper_generic_8bit_conversion_backport_analysis.md
es8375_codec_driver_backport_analysis.md expected_findings_summary.md
hdmi_read_request_analysis.md kernel_analysis_commands.sh
kselftest_harness_teardown_metadata_backport_analysis.md linux-kernel
mediatek_vcodec_h264_backport_analysis.md
pidfd_open_kselftest_harness_compatibility_analysis.md
qualcomm_camss_bus_type_backport_analysis.md
selftests_harness_libatomic_backport_analysis.md
sphinx_version_compatibility_backport_analysis.md
spi_sh_msiof_grpmask_backport_analysis.md
spi_sh_msiof_sitmdr2_sirmdr2_bitfield_conversion_backport_analysis.md
sta2x11_removal_analysis.md test_unused_param.c test_unused_param.o
ti_bridge_encoder_crtc_refactoring_backport_analysis.md
xe_driver_flr_vf_restriction_analysis.md
xe_firmware_per_gt_backport_analysis.md
xe_sriov_logging_backport_analysis.md CPU hotunplug preserves the actual
disabled state when the + amd_display_timing_generator_analysis.md
amd_doorbell_analysis_framework.md
amd_gpu_documentation_backport_analysis.md amd_si_chip_analysis.md
amdgpu_csb_analysis.md amdgpu_mqd_prop_backport_analysis.md
analogix_dp_backport_analysis.md cgroup_rstat_analysis.sh
cgroup_rstat_analysis_guide.md cpuset_comment_fix_backport_analysis.md
csiphy_x1e80100_backport_analysis.md
dcn315_smu_indirect_register_access_analysis.md
drm_dp_helper_dpcd_refactoring_backport_analysis.md
drm_format_helper_generic_8bit_conversion_backport_analysis.md
es8375_codec_driver_backport_analysis.md expected_findings_summary.md
hdmi_read_request_analysis.md kernel_analysis_commands.sh
kselftest_harness_teardown_metadata_backport_analysis.md linux-kernel
mediatek_vcodec_h264_backport_analysis.md
pidfd_open_kselftest_harness_compatibility_analysis.md
qualcomm_camss_bus_type_backport_analysis.md
selftests_harness_libatomic_backport_analysis.md
sphinx_version_compatibility_backport_analysis.md
spi_sh_msiof_grpmask_backport_analysis.md
spi_sh_msiof_sitmdr2_sirmdr2_bitfield_conversion_backport_analysis.md
sta2x11_removal_analysis.md test_unused_param.c test_unused_param.o
ti_bridge_encoder_crtc_refactoring_backport_analysis.md
xe_driver_flr_vf_restriction_analysis.md
xe_firmware_per_gt_backport_analysis.md
xe_sriov_logging_backport_analysis.md CPU comes back online. See
irq_startup_managed(). + linux-kernel/ + desc->depth++; ``` This
preserves the original disable state instead of overwriting it. ### 2.
**New irq_startup_managed() Function**: ```c +void
irq_startup_managed(struct irq_desc *desc) +{ + /bin /bin.usr-is-merged
/boot /dev /etc /home /init /lib /lib.usr-is-merged /lib64 /lost+found
/media /mnt /opt /proc /root /run /sbin /sbin.usr-is-merged /snap /srv
/sys /tmp /usr /var + amd_display_timing_generator_analysis.md
amd_doorbell_analysis_framework.md
amd_gpu_documentation_backport_analysis.md amd_si_chip_analysis.md
amdgpu_csb_analysis.md amdgpu_mqd_prop_backport_analysis.md
analogix_dp_backport_analysis.md cgroup_rstat_analysis.sh
cgroup_rstat_analysis_guide.md cpuset_comment_fix_backport_analysis.md
csiphy_x1e80100_backport_analysis.md
dcn315_smu_indirect_register_access_analysis.md
drm_dp_helper_dpcd_refactoring_backport_analysis.md
drm_format_helper_generic_8bit_conversion_backport_analysis.md
es8375_codec_driver_backport_analysis.md expected_findings_summary.md
hdmi_read_request_analysis.md kernel_analysis_commands.sh
kselftest_harness_teardown_metadata_backport_analysis.md linux-kernel
mediatek_vcodec_h264_backport_analysis.md
pidfd_open_kselftest_harness_compatibility_analysis.md
qualcomm_camss_bus_type_backport_analysis.md
selftests_harness_libatomic_backport_analysis.md
sphinx_version_compatibility_backport_analysis.md
spi_sh_msiof_grpmask_backport_analysis.md
spi_sh_msiof_sitmdr2_sirmdr2_bitfield_conversion_backport_analysis.md
sta2x11_removal_analysis.md test_unused_param.c test_unused_param.o
ti_bridge_encoder_crtc_refactoring_backport_analysis.md
xe_driver_flr_vf_restriction_analysis.md
xe_firmware_per_gt_backport_analysis.md
xe_sriov_logging_backport_analysis.md Only start it up when the disable
depth is 1, so that a disable, +
amd_display_timing_generator_analysis.md
amd_doorbell_analysis_framework.md
amd_gpu_documentation_backport_analysis.md amd_si_chip_analysis.md
amdgpu_csb_analysis.md amdgpu_mqd_prop_backport_analysis.md
analogix_dp_backport_analysis.md cgroup_rstat_analysis.sh
cgroup_rstat_analysis_guide.md cpuset_comment_fix_backport_analysis.md
csiphy_x1e80100_backport_analysis.md
dcn315_smu_indirect_register_access_analysis.md
drm_dp_helper_dpcd_refactoring_backport_analysis.md
drm_format_helper_generic_8bit_conversion_backport_analysis.md
es8375_codec_driver_backport_analysis.md expected_findings_summary.md
hdmi_read_request_analysis.md kernel_analysis_commands.sh
kselftest_harness_teardown_metadata_backport_analysis.md linux-kernel
mediatek_vcodec_h264_backport_analysis.md
pidfd_open_kselftest_harness_compatibility_analysis.md
qualcomm_camss_bus_type_backport_analysis.md
selftests_harness_libatomic_backport_analysis.md
sphinx_version_compatibility_backport_analysis.md
spi_sh_msiof_grpmask_backport_analysis.md
spi_sh_msiof_sitmdr2_sirmdr2_bitfield_conversion_backport_analysis.md
sta2x11_removal_analysis.md test_unused_param.c test_unused_param.o
ti_bridge_encoder_crtc_refactoring_backport_analysis.md
xe_driver_flr_vf_restriction_analysis.md
xe_firmware_per_gt_backport_analysis.md
xe_sriov_logging_backport_analysis.md hotunplug, hotplug sequence does
not end up enabling it during + amd_display_timing_generator_analysis.md
amd_doorbell_analysis_framework.md
amd_gpu_documentation_backport_analysis.md amd_si_chip_analysis.md
amdgpu_csb_analysis.md amdgpu_mqd_prop_backport_analysis.md
analogix_dp_backport_analysis.md cgroup_rstat_analysis.sh
cgroup_rstat_analysis_guide.md cpuset_comment_fix_backport_analysis.md
csiphy_x1e80100_backport_analysis.md
dcn315_smu_indirect_register_access_analysis.md
drm_dp_helper_dpcd_refactoring_backport_analysis.md
drm_format_helper_generic_8bit_conversion_backport_analysis.md
es8375_codec_driver_backport_analysis.md expected_findings_summary.md
hdmi_read_request_analysis.md kernel_analysis_commands.sh
kselftest_harness_teardown_metadata_backport_analysis.md linux-kernel
mediatek_vcodec_h264_backport_analysis.md
pidfd_open_kselftest_harness_compatibility_analysis.md
qualcomm_camss_bus_type_backport_analysis.md
selftests_harness_libatomic_backport_analysis.md
sphinx_version_compatibility_backport_analysis.md
spi_sh_msiof_grpmask_backport_analysis.md
spi_sh_msiof_sitmdr2_sirmdr2_bitfield_conversion_backport_analysis.md
sta2x11_removal_analysis.md test_unused_param.c test_unused_param.o
ti_bridge_encoder_crtc_refactoring_backport_analysis.md
xe_driver_flr_vf_restriction_analysis.md
xe_firmware_per_gt_backport_analysis.md
xe_sriov_logging_backport_analysis.md hotplug unconditionally. + linux-
kernel/ + desc->depth--; + if (!desc->depth) + irq_startup(desc,
IRQ_RESEND, IRQ_START_COND); +} ``` This provides proper depth-aware
startup specifically for managed interrupts. ### 3. **IRQ_STARTUP_ABORT
Case Update**: ```c case IRQ_STARTUP_ABORT: + desc->depth = 1;
irqd_set_managed_shutdown(d); return 0; ``` Ensures consistent depth
handling in abort cases. ### 4. **CPU Hotplug Integration**
(kernel/irq/cpuhotplug.c): ```c - irq_startup(desc, IRQ_RESEND,
IRQ_START_COND); + irq_startup_managed(desc); ``` Uses the new depth-
aware function for managed interrupt restoration. ## Backport
Suitability Assessment **Meets All Stable Criteria:** 1. **Fixes
Important Bug**: ✅ Corrects interrupt enable/disable balance corruption
2. **Small & Contained**: ✅ Minimal changes confined to IRQ subsystem 3.
**Low Regression Risk**: ✅ Only affects managed interrupts during CPU
hotplug 4. **No New Features**: ✅ Pure bug fix, no architectural changes
5. **Critical Subsystem**: ✅ IRQ handling is fundamental to system
stability **Similar Pattern Match**: This closely resembles Similar
Commit #3 and #5 (both marked YES) which also fix managed interrupt CPU
hotplug issues with surgical changes to preserve state across hotplug
events. **Real-World Impact**: The bug can cause: - Server/datacenter
systems using CPU hotplug for power management to experience interrupt
handling issues - Multi-queue device drivers to malfunction after CPU
hotplug cycles - Unbalanced IRQ state leading to system instability The
fix is well-contained, addresses a clear functional bug with user-
visible symptoms, and follows established patterns for stable backports
in the IRQ subsystem.

 kernel/irq/chip.c       | 22 +++++++++++++++++++++-
 kernel/irq/cpuhotplug.c |  2 +-
 kernel/irq/internals.h  |  1 +
 3 files changed, 23 insertions(+), 2 deletions(-)

diff --git a/kernel/irq/chip.c b/kernel/irq/chip.c
index 36cf1b09cc847..ab2bf0de34223 100644
--- a/kernel/irq/chip.c
+++ b/kernel/irq/chip.c
@@ -223,6 +223,19 @@ __irq_startup_managed(struct irq_desc *desc, const struct cpumask *aff,
 		return IRQ_STARTUP_ABORT;
 	return IRQ_STARTUP_MANAGED;
 }
+
+void irq_startup_managed(struct irq_desc *desc)
+{
+	/*
+	 * Only start it up when the disable depth is 1, so that a disable,
+	 * hotunplug, hotplug sequence does not end up enabling it during
+	 * hotplug unconditionally.
+	 */
+	desc->depth--;
+	if (!desc->depth)
+		irq_startup(desc, IRQ_RESEND, IRQ_START_COND);
+}
+
 #else
 static __always_inline int
 __irq_startup_managed(struct irq_desc *desc, const struct cpumask *aff,
@@ -290,6 +303,7 @@ int irq_startup(struct irq_desc *desc, bool resend, bool force)
 			ret = __irq_startup(desc);
 			break;
 		case IRQ_STARTUP_ABORT:
+			desc->depth = 1;
 			irqd_set_managed_shutdown(d);
 			return 0;
 		}
@@ -322,7 +336,13 @@ void irq_shutdown(struct irq_desc *desc)
 {
 	if (irqd_is_started(&desc->irq_data)) {
 		clear_irq_resend(desc);
-		desc->depth = 1;
+		/*
+		 * Increment disable depth, so that a managed shutdown on
+		 * CPU hotunplug preserves the actual disabled state when the
+		 * CPU comes back online. See irq_startup_managed().
+		 */
+		desc->depth++;
+
 		if (desc->irq_data.chip->irq_shutdown) {
 			desc->irq_data.chip->irq_shutdown(&desc->irq_data);
 			irq_state_set_disabled(desc);
diff --git a/kernel/irq/cpuhotplug.c b/kernel/irq/cpuhotplug.c
index 15a7654eff684..3ed5b15927355 100644
--- a/kernel/irq/cpuhotplug.c
+++ b/kernel/irq/cpuhotplug.c
@@ -219,7 +219,7 @@ static void irq_restore_affinity_of_irq(struct irq_desc *desc, unsigned int cpu)
 		return;
 
 	if (irqd_is_managed_and_shutdown(data))
-		irq_startup(desc, IRQ_RESEND, IRQ_START_COND);
+		irq_startup_managed(desc);
 
 	/*
 	 * If the interrupt can only be directed to a single target
diff --git a/kernel/irq/internals.h b/kernel/irq/internals.h
index b0290849c3956..7111747ecb866 100644
--- a/kernel/irq/internals.h
+++ b/kernel/irq/internals.h
@@ -87,6 +87,7 @@ extern void __enable_irq(struct irq_desc *desc);
 extern int irq_activate(struct irq_desc *desc);
 extern int irq_activate_and_startup(struct irq_desc *desc, bool resend);
 extern int irq_startup(struct irq_desc *desc, bool resend, bool force);
+extern void irq_startup_managed(struct irq_desc *desc);
 
 extern void irq_shutdown(struct irq_desc *desc);
 extern void irq_shutdown_and_deactivate(struct irq_desc *desc);
-- 
2.39.5

Re: [PATCH AUTOSEL 6.15 092/110] genirq: Retain disable depth for managed interrupts across CPU hotplug

[Johan Hovold]

On Sun, Jun 01, 2025 at 07:24:14PM -0400, Sasha Levin wrote:
> From: Brian Norris <briannorris@chromium.org>
> 
> [ Upstream commit 788019eb559fd0b365f501467ceafce540e377cc ]
> 
> Affinity-managed interrupts can be shut down and restarted during CPU
> hotunplug/plug. Thereby the interrupt may be left in an unexpected state.
> Specifically:
> 
>  1. Interrupt is affine to CPU N
>  2. disable_irq() -> depth is 1
>  3. CPU N goes offline
>  4. irq_shutdown() -> depth is set to 1 (again)
>  5. CPU N goes online
>  6. irq_startup() -> depth is set to 0 (BUG! driver expects that the interrupt
>     		     	      	        still disabled)
>  7. enable_irq() -> depth underflow / unbalanced enable_irq() warning
> 
> This is only a problem for managed interrupts and CPU hotplug, all other
> cases like request()/free()/request() truly needs to reset a possibly stale
> disable depth value.
> 
> Provide a startup function, which takes the disable depth into account, and
> invoked it for the managed interrupts in the CPU hotplug path.
> 
> This requires to change irq_shutdown() to do a depth increment instead of
> setting it to 1, which allows to retain the disable depth, but is harmless
> for the other code paths using irq_startup(), which will still reset the
> disable depth unconditionally to keep the original correct behaviour.
> 
> A kunit tests will be added separately to cover some of these aspects.
> 
> [ tglx: Massaged changelog ]
> 
> Suggested-by: Thomas Gleixner <tglx@linutronix.de>
> Signed-off-by: Brian Norris <briannorris@chromium.org>
> Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
> Link: https://lore.kernel.org/all/20250514201353.3481400-2-briannorris@chromium.org
> Signed-off-by: Sasha Levin <sashal@kernel.org>

This one breaks suspend of laptops like the Lenovo ThinkPad T14s. Issue
was just reported here by Alex:

	https://lore.kernel.org/lkml/24ec4adc-7c80-49e9-93ee-19908a97ab84@gmail.com/

Please drop from all stable queues for now.

Johan

Re: [PATCH AUTOSEL 6.15 092/110] genirq: Retain disable depth for managed interrupts across CPU hotplug

[Sasha Levin]

On Fri, Jun 06, 2025 at 02:32:29PM +0200, Johan Hovold wrote:
>On Sun, Jun 01, 2025 at 07:24:14PM -0400, Sasha Levin wrote:
>> From: Brian Norris <briannorris@chromium.org>
>>
>> [ Upstream commit 788019eb559fd0b365f501467ceafce540e377cc ]
>>
>> Affinity-managed interrupts can be shut down and restarted during CPU
>> hotunplug/plug. Thereby the interrupt may be left in an unexpected state.
>> Specifically:
>>
>>  1. Interrupt is affine to CPU N
>>  2. disable_irq() -> depth is 1
>>  3. CPU N goes offline
>>  4. irq_shutdown() -> depth is set to 1 (again)
>>  5. CPU N goes online
>>  6. irq_startup() -> depth is set to 0 (BUG! driver expects that the interrupt
>>     		     	      	        still disabled)
>>  7. enable_irq() -> depth underflow / unbalanced enable_irq() warning
>>
>> This is only a problem for managed interrupts and CPU hotplug, all other
>> cases like request()/free()/request() truly needs to reset a possibly stale
>> disable depth value.
>>
>> Provide a startup function, which takes the disable depth into account, and
>> invoked it for the managed interrupts in the CPU hotplug path.
>>
>> This requires to change irq_shutdown() to do a depth increment instead of
>> setting it to 1, which allows to retain the disable depth, but is harmless
>> for the other code paths using irq_startup(), which will still reset the
>> disable depth unconditionally to keep the original correct behaviour.
>>
>> A kunit tests will be added separately to cover some of these aspects.
>>
>> [ tglx: Massaged changelog ]
>>
>> Suggested-by: Thomas Gleixner <tglx@linutronix.de>
>> Signed-off-by: Brian Norris <briannorris@chromium.org>
>> Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
>> Link: https://lore.kernel.org/all/20250514201353.3481400-2-briannorris@chromium.org
>> Signed-off-by: Sasha Levin <sashal@kernel.org>
>
>This one breaks suspend of laptops like the Lenovo ThinkPad T14s. Issue
>was just reported here by Alex:
>
>	https://lore.kernel.org/lkml/24ec4adc-7c80-49e9-93ee-19908a97ab84@gmail.com/
>
>Please drop from all stable queues for now.

Will do, thanks!

-- 
Thanks,
Sasha