[PATCH v5] scsi: ufs: Change the error check sequence for auto hibern8

[PATCH v5] scsi: ufs: Change the error check sequence for auto hibern8

[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

Re: [PATCH v5] scsi: ufs: Change the error check sequence for auto hibern8

[Stanley Jhu]

On 10/27/26 11:10 AM, HOYOUNG SEO wrote:
> 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:
[...]
>  - Move the Signed-off-by above the "---" line and Cc stable.

- The v5 changelog mentions Cc stable, but as far as I can see the
  patch has no "Cc: stable@vger.kernel.org" trailer. Was it meant to
  go right below the Fixes: tag, in the same block as Signed-off-by?

Reviewed-by: Stanley Jhu <stanleyjhu@google.com>