[PATCH] xfstest: log operations properly to generic/001.full

[PATCH] xfstest: log operations properly to generic/001.full

[Ming Chen]

Part of the operations were written to 001.full from the awk script, while the
rest were piped to 001.full using tee.  The resultant 001.full did not reflect
what are really performed.  This makes debugging difficult.

Signed-off-by: Ming Chen <v.mingchen@gmail.com>
---
 tests/generic/001 | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/tests/generic/001 b/tests/generic/001
index 6e5f3dd..e1142ed 100755
--- a/tests/generic/001
+++ b/tests/generic/001
@@ -190,7 +190,7 @@ _mark_iteration()
 #
 _chain()
 {
-    $AWK_PROG -v full_file=$seqres.full -v verify=$verify <$tmp.config '
+    $AWK_PROG -v verify=$verify <$tmp.config '
 BEGIN  { nfile = 0 }
 /^\#/  { next }
        { file[nfile] = $1
@@ -210,7 +210,7 @@ END { srand('$iter')
                printf "if [ ! -f %s ]; then echo \"%s missing!\";
exit; fi\n",file[j],file[j]
                printf "if [ -f %s.0 ]; then echo \"%s.0 already
present!\"; exit; fi\n",file[j],file[j]
                printf "cp %s %s.0 || exit 1\n",file[j],file[j]
-               printf "ls -i %s.0\n", file[j] >full_file;
+               printf "ls -i %s.0\n", file[j];
                total_size += size[j]
                printf "# total size = %d\n", total_size
            }
@@ -219,7 +219,7 @@ END { srand('$iter')
                printf "if [ ! -f %s.%d ]; then echo \"%s.%d
missing!\"; exit; fi\n",file[j],link[j]-1,file[j],link[j]-1
                printf "if [ -f %s.%d ]; then echo \"%s.%d already
present!\"; exit; fi\n",file[j],link[j],file[j],link[j]
                printf "cp %s.%d %s.%d || exit
1\n",file[j],link[j]-1,file[j],link[j]
-               printf "ls -i %s.%d\n", file[j], link[j] >full_file;
+               printf "ls -i %s.%d\n", file[j], link[j];
                total_size += size[j]
                printf "# total size = %d\n", total_size
            }
@@ -231,7 +231,7 @@ END { srand('$iter')
          for (j=0; j<nfile; j++) {
            if (link[j] > 0) {
                printf "mv %s.%d %s.last\n",file[j],link[j]-1,file[j]
-               printf "ls -i %s.last\n", file[j] >full_file;
+               printf "ls -i %s.last\n", file[j];
            }
            for (i=0; i<link[j]-1; i++) {
                printf "rm -f %s.%d\n",file[j],i
-- 
1.8.4.2

_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs

Re: [PATCH] xfstest: log operations properly to generic/001.full

[Dave Chinner]

On Sat, Apr 05, 2014 at 01:52:43AM -0400, Ming Chen wrote:
> Part of the operations were written to 001.full from the awk script, while the
> rest were piped to 001.full using tee.  The resultant 001.full did not reflect
> what are really performed.  This makes debugging difficult.
> 
> Signed-off-by: Ming Chen <v.mingchen@gmail.com>
> ---
>  tests/generic/001 | 8 ++++----
>  1 file changed, 4 insertions(+), 4 deletions(-)
> 
> diff --git a/tests/generic/001 b/tests/generic/001
> index 6e5f3dd..e1142ed 100755
> --- a/tests/generic/001
> +++ b/tests/generic/001
> @@ -190,7 +190,7 @@ _mark_iteration()
>  #
>  _chain()
>  {
> -    $AWK_PROG -v full_file=$seqres.full -v verify=$verify <$tmp.config '
> +    $AWK_PROG -v verify=$verify <$tmp.config '
>  BEGIN  { nfile = 0 }
>  /^\#/  { next }
>         { file[nfile] = $1
> @@ -210,7 +210,7 @@ END { srand('$iter')
>                 printf "if [ ! -f %s ]; then echo \"%s missing!\";
> exit; fi\n",file[j],file[j]
>                 printf "if [ -f %s.0 ]; then echo \"%s.0 already
> present!\"; exit; fi\n",file[j],file[j]

Your mailer program has line wrapped the patch and converted all the
tabs to spaces. I've fixed it up this time, but please read the
Documentation/SubmittingPatches file in the kernel tree to learn how
to prevent this from happening in future.

Cheers,

Dave.
-- 
Dave Chinner
david@fromorbit.com

_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs

Re: [PATCH] xfstest: log operations properly to generic/001.full

[Dave Chinner]

On Sat, Apr 05, 2014 at 01:52:43AM -0400, Ming Chen wrote:
> Part of the operations were written to 001.full from the awk script, while the
> rest were piped to 001.full using tee.  The resultant 001.full did not reflect
> what are really performed.  This makes debugging difficult.
> 
> Signed-off-by: Ming Chen <v.mingchen@gmail.com>

This patch causes generic/001 to fail because it pushes all the
debug into the output file as well as the generic/001.full file.

diff -u tests/generic/001.out /home/dave/src/xfstests-dev/results//generic/001.out.bad
--- tests/generic/001.out       2014-01-20 16:57:33.000000000 +1100
+++ /home/dave/src/xfstests-dev/results//generic/001.out.bad    2014-04-09 09:58:34.000000000 +1000
@@ -1,9 +1,1189 @@
 QA output created by 001
 cleanup
 setup ....................................
-iter 1 chain ... check ....................................
-iter 2 chain ... check ....................................
-iter 3 chain ... check ....................................
-iter 4 chain ... check ....................................
-iter 5 chain ... check ....................................
+iter 1 chain ... 2320100 sub/e.0
+2320101 sub/g.0
+2320102 sub/m00.0
+2320103 sub/b.0
+2320104 sub/d00.0
+2320105 sub/c.0
+2320106 sub/l00.0
+2320107 sub/c.1
.....

I'm going to drop this patch because it breaks a previously working
test....

Cheers,

Dave.
-- 
Dave Chinner
david@fromorbit.com

_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs

Re: [PATCH] xfstest: log operations properly to generic/001.full

[Ming Chen]

On Tue, Apr 8, 2014 at 7:43 PM, Dave Chinner <david@fromorbit.com> wrote:
>
> Your mailer program has line wrapped the patch and converted all the
> tabs to spaces. I've fixed it up this time, but please read the
> Documentation/SubmittingPatches file in the kernel tree to learn how
> to prevent this from happening in future.
>

Sorry, I did not notice. Thanks for the fix and the reference.

Best,
Ming

_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs