Skip to contentSkip to conversation

xfstests generic/263

DISCUSSION J. Bruce Fields 2 participants 6 messages Oct 27, 2020 – Nov 5, 2020

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

Generic/263 is failing whenever client and server both supports READ_PLUS.

I'm not even sure the failure is wrong. The NFS FALLOC operation doesn't support those other other fallocate modes, are they implemented elsewhere in the kernel or libc somehow? Anyway, odd that it would have anything to do with READ_PLUS.

--b.

generic/263 109s ... [failed, exit status 1]- output mismatch (see /root/xfstests-dev/results//generic/263.out.bad)
--- tests/generic/263.out 2019-12-20 17:34:10.493343575 -0500
+++ /root/xfstests-dev/results//generic/263.out.bad 2020-10-27 13:43:41.968835322 -0400
@@ -1,3 +1,2018 @@
QA output created by 263
fsx -N 10000 -o 8192 -l 500000 -r PSIZE -t BSIZE -w BSIZE -Z
-fsx -N 10000 -o 128000 -l 500000 -r PSIZE -t BSIZE -w BSIZE -Z
+Seed set to 1
+main: filesystem does not support fallocate mode FALLOC_FL_KEEP_SIZE, disabling!
+main: filesystem does not support fallocate mode FALLOC_FL_ZERO_RANGE, disabling!
+main: filesystem does not support fallocate mode FALLOC_FL_COLLAPSE_RANGE, disabling!
...
(Run 'diff -u /root/xfstests-dev/tests/generic/263.out /root/xfstests-dev/results//generic/263.out.bad' to see the entire diff)
Ran: generic/263
Failures: generic/263
Failed 1 of 1 tests

#1 lore raw
Quoted 8 lines · J. Bruce Fields

Generic/263 is failing whenever client and server both supports READ_PLUS.

I'm not even sure the failure is wrong. The NFS FALLOC operation doesn't support those other other fallocate modes, are they implemented elsewhere in the kernel or libc somehow? Anyway, odd that it would have anything to do with READ_PLUS.

I just ran xfstests, and I'm seeing this too. The test passes using basic READ on v4.2, so there might be something farther down the log that diff is cutting off. I'll see if anything sticks out to me this week.

Anna

Quoted 19 lines

--b.

generic/263 109s ... [failed, exit status 1]- output mismatch (see /root/xfstests-dev/results//generic/263.out.bad)
--- tests/generic/263.out 2019-12-20 17:34:10.493343575 -0500
+++ /root/xfstests-dev/results//generic/263.out.bad 2020-10-27 13:43:41.968835322 -0400
@@ -1,3 +1,2018 @@
QA output created by 263
fsx -N 10000 -o 8192 -l 500000 -r PSIZE -t BSIZE -w BSIZE -Z
-fsx -N 10000 -o 128000 -l 500000 -r PSIZE -t BSIZE -w BSIZE -Z
+Seed set to 1
+main: filesystem does not support fallocate mode FALLOC_FL_KEEP_SIZE, disabling!
+main: filesystem does not support fallocate mode FALLOC_FL_ZERO_RANGE, disabling!
+main: filesystem does not support fallocate mode FALLOC_FL_COLLAPSE_RANGE, disabling!
...
(Run 'diff -u /root/xfstests-dev/tests/generic/263.out /root/xfstests-dev/results//generic/263.out.bad' to see the entire diff)
Ran: generic/263
Failures: generic/263
Failed 1 of 1 tests

#2 lore raw
Quoted 10 lines · Anna Schumaker
Quoted 8 lines · J. Bruce Fields

Generic/263 is failing whenever client and server both supports READ_PLUS.

I'm not even sure the failure is wrong. The NFS FALLOC operation doesn't support those other other fallocate modes, are they implemented elsewhere in the kernel or libc somehow? Anyway, odd that it would have anything to do with READ_PLUS.

Quoted 4 lines

I just ran xfstests, and I'm seeing this too. The test passes using basic READ on v4.2, so there might be something farther down the log that diff is cutting off. I'll see if anything sticks out to me this week.

Thanks!

Also, wireshark doesn't seem to be parsing READ_PLUS replies correctly. Cc'ing Jorge since he seems to have been the last to touch that code.

--b.

#3 lore raw
Quoted 16 lines · J. Bruce Fields
Quoted 10 lines · Anna Schumaker
Quoted 8 lines · J. Bruce Fields

Generic/263 is failing whenever client and server both supports READ_PLUS.

I'm not even sure the failure is wrong. The NFS FALLOC operation doesn't support those other other fallocate modes, are they implemented elsewhere in the kernel or libc somehow? Anyway, odd that it would have anything to do with READ_PLUS.

Quoted 4 lines

I just ran xfstests, and I'm seeing this too. The test passes using basic READ on v4.2, so there might be something farther down the log that diff is cutting off. I'll see if anything sticks out to me this week.

Quoted 4 lines

Thanks!

Also, wireshark doesn't seem to be parsing READ_PLUS replies correctly. Cc'ing Jorge since he seems to have been the last to touch that code.

Oops, ignore me!

I was actually just running the wrong version of wireshark, with a version built with Jorge's patch it's fine.

--b.

#4 lore raw
In reply to Anna Schumaker
Quoted 10 lines · Anna Schumaker
Quoted 8 lines · J. Bruce Fields

Generic/263 is failing whenever client and server both supports READ_PLUS.

I'm not even sure the failure is wrong. The NFS FALLOC operation doesn't support those other other fallocate modes, are they implemented elsewhere in the kernel or libc somehow? Anyway, odd that it would have anything to do with READ_PLUS.

Quoted 4 lines

I just ran xfstests, and I'm seeing this too. The test passes using basic READ on v4.2, so there might be something farther down the log that diff is cutting off. I'll see if anything sticks out to me this week.

I think 091 is failing from the same cause, by the way. I haven't investigated any more though.

--b.

#5 lore raw
Quoted 12 lines · J. Bruce Fields
Quoted 10 lines · Anna Schumaker
Quoted 8 lines · J. Bruce Fields

Generic/263 is failing whenever client and server both supports READ_PLUS.

I'm not even sure the failure is wrong. The NFS FALLOC operation doesn't support those other other fallocate modes, are they implemented elsewhere in the kernel or libc somehow? Anyway, odd that it would have anything to do with READ_PLUS.

Quoted 7 lines
Quoted 4 lines

I just ran xfstests, and I'm seeing this too. The test passes using basic READ on v4.2, so there might be something farther down the log that diff is cutting off. I'll see if anything sticks out to me this week.

I think 091 is failing from the same cause, by the way. I haven't investigated any more though.

These tests do a combination of directio reads and buffered reads. I don't yet understand why this is a problem with READ_PLUS but works fine for READ

Anna

Quoted 2 lines

--b.