* [PATCH v5] scsi: ufs: Change the error check sequence for auto hibern8
[not found] <CGME20261027021452epcas2p1f71a25c723df847d76cd114ce7afabaa@epcas2p1.samsung.com>
@ 2026-10-27 2:10 ` HOYOUNG SEO
2026-10-27 10:21 ` Stanley Jhu
0 siblings, 1 reply; 2+ messages in thread
From: HOYOUNG SEO @ 2026-10-27 2:10 UTC (permalink / raw)
To: linux-scsi, linux-kernel, alim.akhtar, avri.altman, jejb,
martin.petersen, bvanassche, kwangwon.min, kwmad.kim, cpgs,
h10.kim
Cc: HOYOUNG SEO
Auto-hibern8 enter and exit are performed by the host controller without
a UIC command, and the auto-hibern8 idle timer only runs while no UIC
command is outstanding. An auto-hibern8 error (IS.UHES or IS.UHXS) is
therefore usually reported while hba->active_uic_cmd is NULL.
Since commit fcd8b0450a9a ("scsi: ufs: core: Make ufshcd_uic_cmd_compl()
easier to analyze"), ufshcd_uic_cmd_compl() returns early when
hba->active_uic_cmd is NULL, before ufshcd_is_auto_hibern8_error() is
called. Such auto-hibern8 errors are no longer recorded in hba->errors,
so ufshcd_check_errors() is not called for them and the error handler is
not scheduled.
Check for auto-hibern8 errors before checking for an active UIC command,
as was done before that commit, but keep returning early when there is
no active UIC command so that a NULL cmd is never dereferenced. Only
report the missing UIC command if no auto-hibern8 error was found.
Fixes: fcd8b0450a9a ("scsi: ufs: core: Make ufshcd_uic_cmd_compl() easier to analyze")
Signed-off-by: HOYOUNG SEO <hy50.seo@samsung.com>
---
Changes in v5:
- Fix the unmatched brace in ufshcd_uic_cmd_compl() that broke the build
in v4 (Stanley Jhu, Sashiko, kernel test robot).
- Use a local bool for the ufshcd_is_auto_hibern8_error() result instead
of testing hba->errors (Bart Van Assche).
- Move the Signed-off-by above the "---" line and Cc stable.
- Link to all previous versions.
Changes in v4:
- Changed patch version
Changes in v3:
- Check auto-hibern8 errors before the NULL check again, so that errors
reported while a non-hibernate UIC command is active are still
detected (Sashiko review).
- Return early when no UIC command is active instead of falling
through, to avoid dereferencing a NULL cmd (Sashiko review).
- Add Fixes tag, reword subject and commit message.
Changes in v2:
- Check UFSHCD_UIC_HIBERN8_MASK for auto-hibern8 only when cmd is NULL.
v1: https://lore.kernel.org/linux-scsi/20261021051951.1687479-1-hy50.seo@samsung.com/
v2: https://lore.kernel.org/linux-scsi/20261022074249.3098100-1-hy50.seo@samsung.com/
v3: https://lore.kernel.org/linux-scsi/20261026051851.1845934-1-hy50.seo@samsung.com/
v4: https://lore.kernel.org/linux-scsi/20261026052632.1848929-1-hy50.seo@samsung.com/
drivers/ufs/core/ufshcd.c | 18 +++++++++++++-----
1 file changed, 13 insertions(+), 5 deletions(-)
diff --git a/drivers/ufs/core/ufshcd.c b/drivers/ufs/core/ufshcd.c
index d085d394858f..c08278a2e43e 100644
--- a/drivers/ufs/core/ufshcd.c
+++ b/drivers/ufs/core/ufshcd.c
@@ -5939,18 +5939,26 @@ static irqreturn_t ufshcd_uic_cmd_compl(struct ufs_hba *hba, u32 intr_status)
{
irqreturn_t retval = IRQ_NONE;
struct uic_command *cmd;
+ bool ah8_err;
guard(spinlock_irqsave)(&hba->host->host_lock);
+
+ /*
+ * Auto-hibern8 is performed without a UIC command, so check for
+ * auto-hibern8 errors before checking for an active UIC command.
+ */
+ ah8_err = ufshcd_is_auto_hibern8_error(hba, intr_status);
+ if (ah8_err)
+ hba->errors |= (UFSHCD_UIC_HIBERN8_MASK & intr_status);
+
cmd = hba->active_uic_cmd;
if (!cmd) {
- dev_err(hba->dev,
- "No active UIC command. Maybe a timeout occurred?\n");
+ if (!ah8_err)
+ dev_err(hba->dev,
+ "No active UIC command. Maybe a timeout occurred?\n");
return retval;
}
- if (ufshcd_is_auto_hibern8_error(hba, intr_status))
- hba->errors |= (UFSHCD_UIC_HIBERN8_MASK & intr_status);
-
if (intr_status & UIC_COMMAND_COMPL) {
/*
* Store the UIC command result in the lowest byte of
--
2.34.1
^ permalink raw reply [flat|nested] 2+ messages in thread