Skip to contentSkip to conversation

xfs: Comment out unreachable code within xchk_fscounters()

PATCH In review [email protected] 3 participants 9 messages Jun 6, 2023 – Jun 7, 2023 XFS filesystem

lore
Share & tools
lore

[email protected]

author maintainer
#0 lore raw

PATCH v2 1/1 xfs: Comment out unreachable code within xchk_fscounters()

From: Carlos Maiolino <[email protected]>

Comment the code out so kernel test robot stop complaining about it every single test build.

---
 fs/xfs/scrub/fscounters.c | 7 +++++++
 1 file changed, 7 insertions(+)
fs/xfs/scrub/fscounters.c +7 -0
index e382a35e98d88..228efe0c99be8 100644
@@ -153,6 +153,7 @@ xchk_setup_fscounters(
 	return xchk_trans_alloc(sc, 0);
 }
 
+#if 0
 /*
  * Part 1: Collecting filesystem summary counts.  For each AG, we add its
  * summary counts (total inodes, free inodes, free data blocks) to an incore
@@ -349,6 +350,7 @@ xchk_fscount_count_frextents(
 	return 0;
 }
 #endif /* CONFIG_XFS_RT */
+#endif
 
 /*
  * Part 2: Comparing filesystem summary counters.  All we have to do here is
@@ -422,7 +424,10 @@ xchk_fscounters(
 	struct xfs_mount	*mp = sc->mp;
 	struct xchk_fscounters	*fsc = sc->buf;
 	int64_t			icount, ifree, fdblocks, frextents;
+
+#if 0
 	int			error;
+#endif
 
 	/* Snapshot the percpu counters. */
 	icount = percpu_counter_sum(&mp->m_icount);
@@ -452,6 +457,7 @@ xchk_fscounters(
 	 */
 	return 0;
 
+#if 0
 	/*
 	 * If ifree exceeds icount by more than the minimum variance then
 	 * something's probably wrong with the counters.
@@ -489,4 +495,5 @@ xchk_fscounters(
 		xchk_set_corrupt(sc);
 
 	return 0;
+#endif
 }
Signature · 1 line
2.30.2
Reviewed-by on v2 1/1
#1 lore raw
Quoted 6 lines · On Tue, Jun 06, 2023 at 05:11:22PM +0200, [email protected] wrote:

From: Carlos Maiolino <[email protected]>

Comment the code out so kernel test robot stop complaining about it every single test build.

Thank you,

--D

Quoted 52 lines
---
 fs/xfs/scrub/fscounters.c | 7 +++++++
 1 file changed, 7 insertions(+)
fs/xfs/scrub/fscounters.c +7 -0
index e382a35e98d88..228efe0c99be8 100644
@@ -153,6 +153,7 @@ xchk_setup_fscounters(
 	return xchk_trans_alloc(sc, 0);
 }
 
+#if 0
 /*
  * Part 1: Collecting filesystem summary counts.  For each AG, we add its
  * summary counts (total inodes, free inodes, free data blocks) to an incore
@@ -349,6 +350,7 @@ xchk_fscount_count_frextents(
 	return 0;
 }
 #endif /* CONFIG_XFS_RT */
+#endif
 
 /*
  * Part 2: Comparing filesystem summary counters.  All we have to do here is
@@ -422,7 +424,10 @@ xchk_fscounters(
 	struct xfs_mount	*mp = sc->mp;
 	struct xchk_fscounters	*fsc = sc->buf;
 	int64_t			icount, ifree, fdblocks, frextents;
+
+#if 0
 	int			error;
+#endif
 
 	/* Snapshot the percpu counters. */
 	icount = percpu_counter_sum(&mp->m_icount);
@@ -452,6 +457,7 @@ xchk_fscounters(
 	 */
 	return 0;
 
+#if 0
 	/*
 	 * If ifree exceeds icount by more than the minimum variance then
 	 * something's probably wrong with the counters.
@@ -489,4 +495,5 @@ xchk_fscounters(
 		xchk_set_corrupt(sc);
 
 	return 0;
+#endif
 }
Signature · 1 line
2.30.2
author maintainer
#3 lore raw
Quoted 8 lines · Christoph Hellwig
Quoted 4 lines · On Tue, Jun 06, 2023 at 05:11:22PM +0200, [email protected] wrote:

From: Carlos Maiolino <[email protected]>

Comment the code out so kernel test robot stop complaining about it every single test build.

Err, what? #if 0ing commit coe is a complete no-go. If this code is dead it should be removed.

The code isn't dead, it's temporarily broken. I spoke with Darrick about removing it, but by doing that, later, 'reverting' the patch that removed the broken code, will break the git history (specifically for git blame), and I didn't want to give Darrick extra work by needing to re-add back all this code later when he come back to work on this.
Anyway, just an attempt to quiet built test warning alerts :)
I'm totally fine ^R'ing these emails :)

Signature · 1 line
Carlos
#4 lore raw
Quoted 7 lines · Carlos Maiolino

The code isn't dead, it's temporarily broken. I spoke with Darrick about removing it, but by doing that, later, 'reverting' the patch that removed the broken code, will break the git history (specifically for git blame), and I didn't want to give Darrick extra work by needing to re-add back all this code later when he come back to work on this.
Anyway, just an attempt to quiet built test warning alerts :)
I'm totally fine ^R'ing these emails :)

#if 0 is a realy bad thing. I'd much prefer to remvoe it and re-added it when needed. But even if Darrick insists on just disabling it, you need to add a comment explaining what is going on, because otherwise people will just trip over the complete undocumented #if 0 with a completely meaningless commit message in git-blame. That's how people dealt with code in the early 90s and not now.

author maintainer
#5 lore raw
Quoted 9 lines · Christoph Hellwig
Quoted 7 lines · Carlos Maiolino

The code isn't dead, it's temporarily broken. I spoke with Darrick about removing it, but by doing that, later, 'reverting' the patch that removed the broken code, will break the git history (specifically for git blame), and I didn't want to give Darrick extra work by needing to re-add back all this code later when he come back to work on this.
Anyway, just an attempt to quiet built test warning alerts :)
I'm totally fine ^R'ing these emails :)

Quoted 6 lines

#if 0 is a realy bad thing. I'd much prefer to remvoe it and re-added it when needed. But even if Darrick insists on just disabling it, you need to add a comment explaining what is going on, because otherwise people will just trip over the complete undocumented #if 0 with a completely meaningless commit message in git-blame. That's how people dealt with code in the early 90s and not now.

You are right, I should have added at least some comment on that, I'll wait for Darrick to wake up and see if we deal with it somehow or just leave it as-is.

#6 lore raw
Quoted 17 lines · Carlos Maiolino
Quoted 9 lines · Christoph Hellwig
Quoted 7 lines · Carlos Maiolino

The code isn't dead, it's temporarily broken. I spoke with Darrick about removing it, but by doing that, later, 'reverting' the patch that removed the broken code, will break the git history (specifically for git blame), and I didn't want to give Darrick extra work by needing to re-add back all this code later when he come back to work on this.
Anyway, just an attempt to quiet built test warning alerts :)
I'm totally fine ^R'ing these emails :)

Quoted 6 lines

#if 0 is a realy bad thing. I'd much prefer to remvoe it and re-added it when needed. But even if Darrick insists on just disabling it, you need to add a comment explaining what is going on, because otherwise people will just trip over the complete undocumented #if 0 with a completely meaningless commit message in git-blame. That's how people dealt with code in the early 90s and not now.

Quoted 2 lines

You are right, I should have added at least some comment on that, I'll wait for Darrick to wake up and see if we deal with it somehow or just leave it as-is.

*Someone* please just review the fixes for fscounters.c that I put on the list two weeks ago. The first two patches of the patchset are sufficient to fix this problem.

https://lore.kernel.org/linux-xfs/168506061483.3732954.5178462816967376906.stgit@frogsfrogsfrogs/

--D

#8 lore raw
Quoted 9 lines · Christoph Hellwig
Quoted 5 lines · Darrick J. Wong

*Someone* please just review the fixes for fscounters.c that I put on the list two weeks ago. The first two patches of the patchset are sufficient to fix this problem.

https://lore.kernel.org/linux-xfs/168506061483.3732954.5178462816967376906.stgit@frogsfrogsfrogs/

That sounds like a way better idea indeed. I'll get to it after my meetings.

FYI there's a parallel discussion about the same patch going on fsdevel:

https://lore.kernel.org/linux-fsdevel/20230522234200.GC11598@frogsfrogsfrogs/

--D