Hi all,

After merging the scsi-mkp tree, today's linux-next build (x86_64
allmodconfig) failed like this:

drivers/ufs/core/ufs-mcq.c: In function 'ufshcd_mcq_sq_cleanup':
drivers/ufs/core/ufs-mcq.c:580:9: error: 'rtc' undeclared (first use in this function)
  580 |         rtc = FIELD_GET(SQ_ICU_ERR_CODE_MASK, readl(reg));
      |         ^~~
drivers/ufs/core/ufs-mcq.c:580:9: note: each undeclared identifier is reported only once for each function it appears in

Caused by commit

  bedea6f472ab ("Merge branch 'for-next' of git://git.kernel.org/pub/scm/linux/kernel/git/mkp/scsi.git")

The automatic merge resolution didn't work out right.

I have applied the following fix up patch (that may be needed when the
scsi and scsi-mkp trees are merged): 

From: Stephen Rothwell <sfr@canb.auug.org.au>
Date: Thu, 7 Nov 2024 21:21:01 +1100
Subject: [PATCH] bad automatic merge fixup for scsi-mkp merge

Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
---
 drivers/ufs/core/ufs-mcq.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/ufs/core/ufs-mcq.c b/drivers/ufs/core/ufs-mcq.c
index fa25e9ac2804..18ca95e5b68c 100644
--- a/drivers/ufs/core/ufs-mcq.c
+++ b/drivers/ufs/core/ufs-mcq.c
@@ -539,7 +539,7 @@ int ufshcd_mcq_sq_cleanup(struct ufs_hba *hba, int task_tag)
 	struct scsi_cmnd *cmd = lrbp->cmd;
 	struct ufs_hw_queue *hwq;
 	void __iomem *reg, *opr_sqd_base;
-	u32 nexus, id, val;
+	u32 nexus, id, val, rtc;
 	int err;
 
 	if (hba->quirks & UFSHCD_QUIRK_MCQ_BROKEN_RTC)
-- 
2.45.2

-- 
Cheers,
Stephen Rothwell

Hi Stephen!

> After merging the scsi-mkp tree, today's linux-next build (x86_64
> allmodconfig) failed like this:
>
> drivers/ufs/core/ufs-mcq.c: In function 'ufshcd_mcq_sq_cleanup':
> drivers/ufs/core/ufs-mcq.c:580:9: error: 'rtc' undeclared (first use in this function)
>   580 |         rtc = FIELD_GET(SQ_ICU_ERR_CODE_MASK, readl(reg));
>       |         ^~~
> drivers/ufs/core/ufs-mcq.c:580:9: note: each undeclared identifier is reported only once for each function it appears in

Hrm, I believe I already addressed this conflict in the SCSI tree.

Maybe James' repo has stale bits?

-- 
Martin K. Petersen	Oracle Linux Engineering

On Thu, 2024-11-07 at 15:47 -0500, Martin K. Petersen wrote:
> 
> Hi Stephen!
> 
> > After merging the scsi-mkp tree, today's linux-next build (x86_64
> > allmodconfig) failed like this:
> > 
> > drivers/ufs/core/ufs-mcq.c: In function 'ufshcd_mcq_sq_cleanup':
> > drivers/ufs/core/ufs-mcq.c:580:9: error: 'rtc' undeclared (first
> > use in this function)
> >   580 |         rtc = FIELD_GET(SQ_ICU_ERR_CODE_MASK, readl(reg));
> >       |         ^~~
> > drivers/ufs/core/ufs-mcq.c:580:9: note: each undeclared identifier
> > is reported only once for each function it appears in
> 
> Hrm, I believe I already addressed this conflict in the SCSI tree.
> 
> Maybe James' repo has stale bits?

No, my tree builds ... or at least the ufs-mcq.c part of it (I checked
after I got the merge conflict ... although only with the default
configuration).  I'm still worried about the resolution I flagged on
the list, though.

Regards,

James

James,

> No, my tree builds ... or at least the ufs-mcq.c part of it (I checked
> after I got the merge conflict ... although only with the default
> configuration).

I am not questioning that your tree builds. But your for-next branch
contains UFS code not present in the SCSI tree, effectively reverting my
conflict resolution.

-- 
Martin K. Petersen	Oracle Linux Engineering

On Thu, 2024-11-07 at 16:31 -0500, Martin K. Petersen wrote:
> 
> James,
> 
> > No, my tree builds ... or at least the ufs-mcq.c part of it (I
> > checked
> > after I got the merge conflict ... although only with the default
> > configuration).
> 
> I am not questioning that your tree builds. But your for-next branch
> contains UFS code not present in the SCSI tree, effectively reverting
> my conflict resolution.

OK, I figured it out.  We both did the conflict resolution for "scsi:
ufs: core: Fix another deadlock during RTC update" slightly
differently.  I kept the rtc variable introduced in that commit and you
removed it leading to the conflict.  Since it's only in a print, I
don't think it matters, so I followed your resolution.

James