Skip to contentSkip to conversation

xfsprogs 3.1.12 and 3.2.0 releases?

DISCUSSION Arkadiusz Miśkiewicz 7 participants 32 messages Jan 20, 2014 – Apr 8, 2014

lore
Share & tools
lore

Arkadiusz Miśkiewicz

author
#0 lore raw

Hello.

What's needed for both 3.1.12 and 3.2.0 releases to happen?

If 3.2.0 is to be released soon then nice. Dave, AFAIK it is waiting for you and xfs_db crc work only?

If 3.2.0 is far away then it would be nice to see 3.1.12 with bugfixes only (for those paranoid about new stuff) but 3.1.13 very soon after that with backported features like project quota support, parallelism, performance improvements.

Obviously I hope 3.2.0 is soon and there is no need to waste effort on 3.1.x.

Signature · 6 lines
Arkadiusz Miśkiewicz, arekm / maven.pl

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs
#1 lore raw
Quoted 7 lines · Arkadiusz Miśkiewicz

Hello.

What's needed for both 3.1.12 and 3.2.0 releases to happen?

If 3.2.0 is to be released soon then nice. Dave, AFAIK it is waiting for you and xfs_db crc work only?

xfs_repair work, actually. xfs_db is pretty much complete.

Quoted 4 lines

If 3.2.0 is far away then it would be nice to see 3.1.12 with bugfixes only (for those paranoid about new stuff) but 3.1.13 very soon after that with backported features like project quota support, parallelism, performance improvements.

You're not going to get a 3.1.x release with anything other than bug fixes in it. Indeed, separate project quota inode support requires version 5 superblocks, so that's not ever going to be backported to the 3.1.x series...

Quoted 1 line

Obviously I hope 3.2.0 is soon and there is no need to waste effort on 3.1.x.

The problem for me right now is that there are only so many hours in the day....

Cheers,

Dave.

Signature · 7 lines
Dave Chinner
[email protected]

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs
author
#2 lore raw
Quoted 9 lines · Dave Chinner
Quoted 6 lines · Arkadiusz Miśkiewicz

Hello.

What's needed for both 3.1.12 and 3.2.0 releases to happen?

If 3.2.0 is to be released soon then nice. Dave, AFAIK it is waiting for you and xfs_db crc work only?

xfs_repair work, actually. xfs_db is pretty much complete.

[...]

Quoted 5 lines
Quoted 2 lines

Obviously I hope 3.2.0 is soon and there is no need to waste effort on 3.1.x.

The problem for me right now is that there are only so many hours in the day....

Could 3.2.0 be released without full xfs_db crc support then? And defer xfs_db to 3.2.1 or even later. Especially you say "xfs_db is pretty much complete".

Right now we have a situation where some features exist in kernel (as stable feature - separate project quota) that are not supported by userspace. Some fixes (growfs thing/superblock garbage) missing etc. Improvements (performance/paralellism) waiting.

ps. I assume that someone else does the actual release, so no need for more hours/day for you in such case.

Quoted 2 lines

Cheers,
Dave.

Signature · 6 lines
Arkadiusz Miśkiewicz, arekm / maven.pl

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs
#3 lore raw
Quoted 13 lines · Arkadiusz Miśkiewicz
Quoted 9 lines · Dave Chinner
Quoted 6 lines · Arkadiusz Miśkiewicz

Hello.

What's needed for both 3.1.12 and 3.2.0 releases to happen?

If 3.2.0 is to be released soon then nice. Dave, AFAIK it is waiting for you and xfs_db crc work only?

xfs_repair work, actually. xfs_db is pretty much complete.

[...]

Quoted 8 lines
Quoted 5 lines
Quoted 2 lines

Obviously I hope 3.2.0 is soon and there is no need to waste effort on 3.1.x.

The problem for me right now is that there are only so many hours in the day....

Could 3.2.0 be released without full xfs_db crc support then? And defer xfs_db to 3.2.1 or even later. Especially you say "xfs_db is pretty much complete".

* release without full /xfs_repair/ support, not xfs_db.

No, I don't think so. releasing without full xfs_db support isn't an issue as it's not critical to users. OTOH, xfs_repair is critical to users.

Quoted 2 lines

Right now we have a situation where some features exist in kernel (as stable feature - separate project quota) that are not supported by userspace.

Separate project quota is v5 superblocks only, which are still marked EXPERIMENTAL. For everything non-v5 superblock, v3.1.11 works fine (modulo bugs).

Quoted 3 lines

Some
fixes (growfs thing/superblock garbage) missing etc. Improvements (performance/paralellism) waiting.

Nobody is going to push new features back to 3.1.x - we don't have neough developer resources available to do that sort of work. That's pretty obvious, given we are even having this conversation.

Quoted 2 lines

ps. I assume that someone else does the actual release, so no need for more hours/day for you in such case.

Well, that's always been the plan since a 3.1.12 release was proposed 3 months ago. How well has that plan been working out so far?

Cheers,

Dave.

Signature · 7 lines
Dave Chinner
[email protected]

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs
#4 lore raw
Quoted 3 lines · Dave Chinner

Well, that's always been the plan since a 3.1.12 release was proposed 3 months ago. How well has that plan been working out so far?

The sad part is that I would have had the time 3 month ago, but right now it's hard. Still trying to get it done, though.

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs

#5 lore raw
Quoted 7 lines · Christoph Hellwig
Quoted 3 lines · Dave Chinner

Well, that's always been the plan since a 3.1.12 release was proposed 3 months ago. How well has that plan been working out so far?

The sad part is that I would have had the time 3 month ago, but right now it's hard. Still trying to get it done, though.

If Arkadiusz is willing to provide a list of commits to include it shouldn't be a problem to get this done. I believe he indicated a willingness to help on IRC earlier today.

-Ben

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs

author
#6 lore raw
Quoted 6 lines · Ben Myers
Quoted 5 lines · Christoph Hellwig
Quoted 3 lines · Dave Chinner

Well, that's always been the plan since a 3.1.12 release was proposed 3 months ago. How well has that plan been working out so far?

Quoted 6 lines
Quoted 2 lines

The sad part is that I would have had the time 3 month ago, but right now it's hard. Still trying to get it done, though.

If Arkadiusz is willing to provide a list of commits to include it shouldn't be a problem to get this done. I believe he indicated a willingness to help on IRC earlier today.

By looking into git log these look like small fixes for 3.1.x. Could someone else recheck?

108e985bc596a1f1a33d31a348fdb430213fb8c0 xfs_fsr: file reads should be O_DIRECT 46f34a259939d8e815da1a5a5374d39153ed57e3 xfs_metadump: manpage fix regarding frozen fs 1b6bf714a7179942b8523cc966124249b74381da mkfs.xfs: fix protofile name create block reservation 504dbe46549d462e79565514d86ce69b74b96893 xfsprogs: avoid double-free in xfs_attr_node_addname a134873e2903837bdbeab261b6d0ceee2fd499ba xfsprogs: fix buffer leak in xlog_print_find_oldest 4623d1041dd34d73aeab4b51002fc9ca6d543415 xfsprogs: free bp in xlog_find_zeroed() error path dc93954aa6f13963bcc87fc00ce55a4745dc7b93 xfsprogs:free bp in xlog_find_tail() error path aba29588f81dad52129037c15fe584ec89b36a4f xfsprogs: fix agcnts leak in xfs_repair's scan_ags 312592defb8b6051389a56a5c780819b4239dab4 xfsprogs: fix unint var in repair phase6 3ac87fbf6ac29494dc46134323733551aac1737c xfsprogs: fix inode crash in xfs_repair e7c05095f5baa9cd2e35a6de03d7dd9f51dd3910 xfsprogs: fix Out-of-bounds access in repair/dinode.c cbd7508db4c9597889ad98d5f027542002e0e57c xfs_repair: zero out unused parts of superblocks [we had real issue that this fixes] e6efb967e61a366dbe877f34e220e32866e7db42 xfsprogs: fix Out-of-bounds access in repair/dinode.c e49f30a76f5e910092bd285ebef3f7641a09ec2e xfsprogs: avoid array overflow in pf_batch_read() 70ee4153f4c998774adcee1e576d3c1f305b92d0 xfsprogs: fix potential memory leak in repare/sb.c 548c2e3e5b123266b4b89cc81bedd113442570b8 [v3, 1/2] xfsprogs: fix potential memory leak in verify_set_primary_sb() ae181820d03e19f145dbd058c0f8fa59fa18468c xfsprogs: fix return value of verify_set_primary_sb() 050a7f1f79314470c18e9d11c28daf75dd959830 xfsprogs: handle symlinks etc in fs_table_initialise_mounts() 773d6f6e3800dce5d27af687ebd3ff323dcdfd72 xfsprogs: config.guess/config.sub out of date for arm64 7f2d6b811755b6b91f18aa5bd9d5980848a81267 xfs_repair: avoid segfault if reporting progress early in repair 44dae5e6804408b4123a916a2738b73e21d8c61e xfs_repair: test for bad level in dir2 node 14e36e34ed975b1ffbbcbc18a7743a3a4859be3d xfsprogs: fix resource leak in longform_dir2_rebuild() 12a48f5d8a06a8ca21b6221e94318e4922f0316c xfsprogs: remove xfs_check references from fsck.xfs script & manpage 6aa3d87bc45348dc0948ae0cea57bf3033d64694 xfs: don't emit corruption noise on fs probes 9aa5711629b47642bb5b688a6a1410d223456fc8 libxfs: fix root inode handling inconsistencies 3a19fb7dce9d570e78deaf5c26c0ab8a4a5bef67 libxfs: stop caching inode structures 61510437c627b529feb95ebffddd73df5ed5b104 repair: prefetching is turned off unnecessarily 0cce4aa198f0470817bedb3781ea5b6955e43076 repair: Increase default repair parallelism on large filesystems 1adfe5c6296d3ea6c182f31a6728fc94af9146f7 xfs_fsr: fix SWAPEXT failures under selinux 79e106f05f0fbffb89ab5401fb97af5391c6291f xfs_repair: correct docs for "-t" units 03091fc253963891c728cf362c0f9700294b6dfd xfsprogs: suggest "-d" option for repair of RO mount 3ae81520e91945454f35fa75833a45b5b000fe33 xfs_repair: suggest reboot after dangerous repair 4e83ac7b7416edb52815c6e5f4ca64dadfa5f974 xfs_metadump: Make -F (force) optional

Not sure about these:

1ba28b64e69ce3a7989df88f71a5cb608b1c71e3 xfs_repair: drop buffer reference on symlink error 4e503735fd407e2e61295e6af6ec47af4693bc95 xfs_repair: fix btree block magic number mapping 4fbebf374ccc178f1fedcf2d8a43339031f8dbb4 libxfs: fix dir3 freespace block corruption 494434d7fb79840ba113ecd7fb1ac3ae20e0f569 xfs: Add read-only support for dirent filetype field 906b762f55333968321062642a0b90feac1fdffb xfs: Add write support for dirent filetype field 3beed08eb22f56b384d2028541ccb41284ff9751 xfsprogs: add dtype support to mkfs and db 1acc538540ce22e16bb55ca573691070a8d375db xfsprogs: initialize filetype for xfs_name_dot 41315687d9db9b50876401e7b0ee20dd77cfc712 xfs: dirent dtype presence is dependent on directory magic numbers 68774b900e0c8368342cb12f649572a86ef2f6e4 xfsprogs: initialize filetype for lost+found creation 2cca1dbd1c3e12d270a9baa5f85e548e8a5a2125 xfs_repair: add d_type when moving files to lost+found 87e343bd0937e5bb75dd8bc46ba388b6f8c6552b xfsprog: add xfs sb v4 support for dirent filetype field 6f700630b06a2ce15aebe8608b2c5877002299d6 xfsprog: add dirent filetype information for xfs_info 42737f1ad16213a3dab1756c9fffb494db8ef27e xfs_progs: add dirent filetype to xfs_db version 4eb02d95b7e081b510a7015609f01385aab229a9 xfsprog: add mkfs.xfs sb v4 support for dirent filetype field 12864fd992dd5d6bc3c089aeb6422c8d235a28f0 xfs: remove newlines from strings passed to __xfs_printk 3e23516ae60421652fd41354307a6a5181d401eb xfs: fix the wrong new_size/rnew_size at xfs_iext_realloc_direct() c9522f4d8790ecd61c4e74746b607787485f2027 xfs: fix node forward in xfs_node_toosmall 839dac7f06d54600b3092a7ad9cb903315a27f97 libxfs: work around do_div() not handling 32 bit numerators ea4a8de1e135905e240a54f192daf23275a7e8a8 xfs_repair: fix process_bmbt_reclist_int

Quoted 1 line

-Ben

Signature · 6 lines
Arkadiusz Miśkiewicz, arekm / maven.pl

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs
#7 lore raw
Quoted 16 lines · Arkadiusz Miśkiewicz
Quoted 6 lines · Ben Myers
Quoted 5 lines · Christoph Hellwig
Quoted 3 lines · Dave Chinner

Well, that's always been the plan since a 3.1.12 release was proposed 3 months ago. How well has that plan been working out so far?

Quoted 6 lines
Quoted 2 lines

The sad part is that I would have had the time 3 month ago, but right now it's hard. Still trying to get it done, though.

If Arkadiusz is willing to provide a list of commits to include it shouldn't be a problem to get this done. I believe he indicated a willingness to help on IRC earlier today.

By looking into git log these look like small fixes for 3.1.x. Could someone else recheck?

.....

Quoted 3 lines

3a19fb7dce9d570e78deaf5c26c0ab8a4a5bef67 libxfs: stop caching inode structures 61510437c627b529feb95ebffddd73df5ed5b104 repair: prefetching is turned off unnecessarily 0cce4aa198f0470817bedb3781ea5b6955e43076 repair: Increase default repair parallelism on large filesystems

I think these three are not a great idea for 3.1.x. They result in significant changes of behaviour that can result in increased memory consumption by default. I'd prefer that we don't make changes to the default behaviour in a bug-fix only point release.

.....

Quoted 16 lines

Not sure about these:

1ba28b64e69ce3a7989df88f71a5cb608b1c71e3 xfs_repair: drop buffer reference on symlink error 4e503735fd407e2e61295e6af6ec47af4693bc95 xfs_repair: fix btree block magic number mapping 4fbebf374ccc178f1fedcf2d8a43339031f8dbb4 libxfs: fix dir3 freespace block corruption 494434d7fb79840ba113ecd7fb1ac3ae20e0f569 xfs: Add read-only support for dirent filetype field 906b762f55333968321062642a0b90feac1fdffb xfs: Add write support for dirent filetype field 3beed08eb22f56b384d2028541ccb41284ff9751 xfsprogs: add dtype support to mkfs and db 1acc538540ce22e16bb55ca573691070a8d375db xfsprogs: initialize filetype for xfs_name_dot 41315687d9db9b50876401e7b0ee20dd77cfc712 xfs: dirent dtype presence is dependent on directory magic numbers 68774b900e0c8368342cb12f649572a86ef2f6e4 xfsprogs: initialize filetype for lost+found creation 2cca1dbd1c3e12d270a9baa5f85e548e8a5a2125 xfs_repair: add d_type when moving files to lost+found 87e343bd0937e5bb75dd8bc46ba388b6f8c6552b xfsprog: add xfs sb v4 support for dirent filetype field 6f700630b06a2ce15aebe8608b2c5877002299d6 xfsprog: add dirent filetype information for xfs_info 42737f1ad16213a3dab1756c9fffb494db8ef27e xfs_progs: add dirent filetype to xfs_db version 4eb02d95b7e081b510a7015609f01385aab229a9 xfsprog: add mkfs.xfs sb v4 support for dirent filetype field

I'd say not to the dirent filetype changes. They require the CRC format enabled directory code and that basically involves backporting the entire 3.2-alpha code base to support it. Otherwise a complete re-implementation in needed for 3.1.x, not a backport.

Cheers,

Dave.

Signature · 7 lines
Dave Chinner
[email protected]

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs
#8 lore raw
Quoted 45 lines · Dave Chinner
Quoted 16 lines · Arkadiusz Miśkiewicz
Quoted 6 lines · Ben Myers
Quoted 5 lines · Christoph Hellwig
Quoted 3 lines · Dave Chinner

Well, that's always been the plan since a 3.1.12 release was proposed 3 months ago. How well has that plan been working out so far?

Quoted 6 lines
Quoted 2 lines

The sad part is that I would have had the time 3 month ago, but right now it's hard. Still trying to get it done, though.

If Arkadiusz is willing to provide a list of commits to include it shouldn't be a problem to get this done. I believe he indicated a willingness to help on IRC earlier today.

By looking into git log these look like small fixes for 3.1.x. Could someone else recheck?

.....

Quoted 3 lines

3a19fb7dce9d570e78deaf5c26c0ab8a4a5bef67 libxfs: stop caching inode structures 61510437c627b529feb95ebffddd73df5ed5b104 repair: prefetching is turned off unnecessarily 0cce4aa198f0470817bedb3781ea5b6955e43076 repair: Increase default repair parallelism on large filesystems

I think these three are not a great idea for 3.1.x. They result in significant changes of behaviour that can result in increased memory consumption by default. I'd prefer that we don't make changes to the default behaviour in a bug-fix only point release.

.....

Quoted 16 lines

Not sure about these:

1ba28b64e69ce3a7989df88f71a5cb608b1c71e3 xfs_repair: drop buffer reference on symlink error 4e503735fd407e2e61295e6af6ec47af4693bc95 xfs_repair: fix btree block magic number mapping 4fbebf374ccc178f1fedcf2d8a43339031f8dbb4 libxfs: fix dir3 freespace block corruption 494434d7fb79840ba113ecd7fb1ac3ae20e0f569 xfs: Add read-only support for dirent filetype field 906b762f55333968321062642a0b90feac1fdffb xfs: Add write support for dirent filetype field 3beed08eb22f56b384d2028541ccb41284ff9751 xfsprogs: add dtype support to mkfs and db 1acc538540ce22e16bb55ca573691070a8d375db xfsprogs: initialize filetype for xfs_name_dot 41315687d9db9b50876401e7b0ee20dd77cfc712 xfs: dirent dtype presence is dependent on directory magic numbers 68774b900e0c8368342cb12f649572a86ef2f6e4 xfsprogs: initialize filetype for lost+found creation 2cca1dbd1c3e12d270a9baa5f85e548e8a5a2125 xfs_repair: add d_type when moving files to lost+found 87e343bd0937e5bb75dd8bc46ba388b6f8c6552b xfsprog: add xfs sb v4 support for dirent filetype field 6f700630b06a2ce15aebe8608b2c5877002299d6 xfsprog: add dirent filetype information for xfs_info 42737f1ad16213a3dab1756c9fffb494db8ef27e xfs_progs: add dirent filetype to xfs_db version 4eb02d95b7e081b510a7015609f01385aab229a9 xfsprog: add mkfs.xfs sb v4 support for dirent filetype field

Quoted 8 lines

I'd say not to the dirent filetype changes. They require the CRC format enabled directory code and that basically involves backporting the entire 3.2-alpha code base to support it. Otherwise a complete re-implementation in needed for 3.1.x, not a backport.

Cheers,

Dave.

Thanks for making the list. I agree with Dave above about 3 feature patches and the directory file type changes.

The following patch that is in your "not sure" list looks like a good bug fix that is not CRC related:

ba28b64e69ce3a7989df88f71a5cb608b1c71e3 xfs_repair: drop buffer reference on symlink error

--Mark.

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs

#9 lore raw

On 01/22/14 08:38, Mark Tinguely wrote:
...

Quoted 9 lines

Thanks for making the list. I agree with Dave above about 3 feature patches and the directory file type changes.

The following patch that is in your "not sure" list looks like a good bug fix that is not CRC related:

ba28b64e69ce3a7989df88f71a5cb608b1c71e3 xfs_repair: drop buffer reference on symlink error

Take that back, if (fsbno == NULLDFSBNO) then bp will be NULL also. So we do not want to do the libxfs_putbuf()... this patch is not appropriate for 3.1.12.

--Mark.

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs

author
#10 lore raw
In reply to Dave Chinner
Quoted 12 lines · Dave Chinner
Quoted 2 lines · Arkadiusz Miśkiewicz

By looking into git log these look like small fixes for 3.1.x. Could someone else recheck?

.....

Quoted 4 lines

3a19fb7dce9d570e78deaf5c26c0ab8a4a5bef67 libxfs: stop caching inode structures 61510437c627b529feb95ebffddd73df5ed5b104 repair: prefetching is turned off unnecessarily 0cce4aa198f0470817bedb3781ea5b6955e43076
repair: Increase default repair parallelism on large filesystems

Quoted 4 lines

I think these three are not a great idea for 3.1.x. They result in significant changes of behaviour that can result in increased memory consumption by default. I'd prefer that we don't make changes to the default behaviour in a bug-fix only point release.

3a19fb7dce9d570e78deaf5c26c0ab8a4a5bef67 libxfs: stop caching inode

looks like wasting memory fix to me but ok.

Quoted 1 line

Dave.

Signature · 6 lines
Arkadiusz Miśkiewicz, arekm / maven.pl

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs
#11 lore raw
Quoted 13 lines · Arkadiusz Miśkiewicz
Quoted 12 lines · Dave Chinner
Quoted 2 lines · Arkadiusz Miśkiewicz

By looking into git log these look like small fixes for 3.1.x. Could someone else recheck?

.....

Quoted 4 lines

3a19fb7dce9d570e78deaf5c26c0ab8a4a5bef67 libxfs: stop caching inode structures 61510437c627b529feb95ebffddd73df5ed5b104 repair: prefetching is turned off unnecessarily 0cce4aa198f0470817bedb3781ea5b6955e43076
repair: Increase default repair parallelism on large filesystems

Quoted 8 lines
Quoted 4 lines

I think these three are not a great idea for 3.1.x. They result in significant changes of behaviour that can result in increased memory consumption by default. I'd prefer that we don't make changes to the default behaviour in a bug-fix only point release.

3a19fb7dce9d570e78deaf5c26c0ab8a4a5bef67 libxfs: stop caching inode

looks like wasting memory fix to me but ok.

The last paragraph ofi the commit message points out the problems that can occur with memory consumption. For sparse inode populations, the buffer cache can consume much more memory than the inode cache did, and that can cause OOM problems with repair...

As it is, it is a significant change of behaviour, so for a lightweight bug-fix only release I feel that it is out-of-scope...

Cheers,

Dave.

Signature · 7 lines
Dave Chinner
[email protected]

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs
#12 lore raw
Quoted 28 lines · Dave Chinner
Quoted 13 lines · Arkadiusz Miśkiewicz
Quoted 12 lines · Dave Chinner
Quoted 2 lines · Arkadiusz Miśkiewicz

By looking into git log these look like small fixes for 3.1.x. Could someone else recheck?

.....

Quoted 4 lines

3a19fb7dce9d570e78deaf5c26c0ab8a4a5bef67 libxfs: stop caching inode structures 61510437c627b529feb95ebffddd73df5ed5b104 repair: prefetching is turned off unnecessarily 0cce4aa198f0470817bedb3781ea5b6955e43076
repair: Increase default repair parallelism on large filesystems

Quoted 8 lines
Quoted 4 lines

I think these three are not a great idea for 3.1.x. They result in significant changes of behaviour that can result in increased memory consumption by default. I'd prefer that we don't make changes to the default behaviour in a bug-fix only point release.

3a19fb7dce9d570e78deaf5c26c0ab8a4a5bef67 libxfs: stop caching inode

looks like wasting memory fix to me but ok.

The last paragraph ofi the commit message points out the problems that can occur with memory consumption. For sparse inode populations, the buffer cache can consume much more memory than the inode cache did, and that can cause OOM problems with repair...

Quoted 7 lines

As it is, it is a significant change of behaviour, so for a lightweight bug-fix only release I feel that it is out-of-scope...

Cheers,

Dave.

I'll get this put together by Monday unless someone else wants to do it.

--Rich

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs

#13 lore raw
In reply to Ben Myers
Quoted 3 lines · Ben Myers

If Arkadiusz is willing to provide a list of commits to include it shouldn't be a problem to get this done. I believe he indicated a willingness to help on IRC earlier today.

I already provided a list a while ago, although that would have grown a bit by now.

But applying the patches isn't really the hard part, the time consuming bit is to provide proper QA.

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs

#14 lore raw

[ANNOUNCE] xfsprogs for-3.1.12 branch has been updated to 40c65a7

In reply to Arkadiusz Miśkiewicz

Arkadiusz,

Thanks for compiling the list.

The for-3.1.12 Branch created from "a547280... 3.1.11 release"

If there are others you want committed, please let me know.

Thanks
--Rich

The following have been committed from the candidate list provided by Arkadiusz:

40c65a7... xfs_metadump: Make -F (force) optional
ec3a1f1... xfs_repair: suggest reboot after dangerous repair
8660952... xfsprogs: suggest "-d" option for repair of RO mount
5a8206c... xfs_repair: correct docs for "-t" units
0d5c444... xfs_fsr: fix SWAPEXT failures under selinux
f3d3dae... libxfs: fix root inode handling inconsistencies
cf507b6... xfsprogs: remove xfs_check references from fsck.xfs script & manpage
6e2f2d1... xfsprogs: fix resource leak in longform_dir2_rebuild()
ecde3f9... xfs_repair: test for bad level in dir2 node
d299df7... xfs_repair: avoid segfault if reporting progress early in repair
e660c2e... Sent: Friday, October 11, 2013 12:38:11 AM Subject: Bug#725971: xfsprogs: config.guess/config.sub out of date for arm64 5e1c7ac... xfsprogs: handle symlinks etc in fs_table_initialise_mounts()
aba6204... xfsprogs: fix return value of verify_set_primary_sb()
a2ea195... xfsprogs: fix potential memory leak in verify_set_primary_sb()
3e53939... xfsprogs: fix potential memory leak in repare/sb.c
820b8c2... xfsprogs: avoid array overflow in pf_batch_read()
2a1dcdb... xfs_repair: zero out unused parts of superblocks
982508b... xfsprogs: fix inode crash in xfs_repair
b8890d7... xfsprogs: fix unint var in repair phase6
7d710b9... xfsprogs: fix agcnts leak in xfs_repair's scan_ags
291c159... xfsprogs:free bp in xlog_find_tail() error path
40b9d95... xfsprogs: free bp in xlog_find_zeroed() error path
a9b4757... xfsprogs: fix buffer leak in xlog_print_find_oldest
32aae5b... xfsprogs: avoid double-free in xfs_attr_node_addname
243cf82... mkfs.xfs: fix protofile name create block reservation
b31a23d... xfs_metadump: manpage fix regarding frozen fs
f8b80a3... xfs_fsr: file reads should be O_DIRECT

The following commits were not applicable.
e7c05095, e6efb967 6aa3d87b

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs

#15 lore raw

Re: [ANNOUNCE] xfsprogs for-3.1.12 branch has been updated to 40c65a7

Quoted 7 lines · Rich Johnston

Arkadiusz,

Thanks for compiling the list.

The for-3.1.12 Branch created from "a547280... 3.1.11 release"

If there are others you want committed, please let me know.

What testing have you done on the branch?

Quoted 5 lines

40c65a7... xfs_metadump: Make -F (force) optional ec3a1f1... xfs_repair: suggest reboot after dangerous repair 8660952... xfsprogs: suggest "-d" option for repair of RO mount 5a8206c... xfs_repair: correct docs for "-t" units 0d5c444... xfs_fsr: fix SWAPEXT failures under selinux

Quoted 8 lines

f3d3dae... libxfs: fix root inode handling inconsistencies cf507b6... xfsprogs: remove xfs_check references from fsck.xfs script & manpage
6e2f2d1... xfsprogs: fix resource leak in longform_dir2_rebuild() ecde3f9... xfs_repair: test for bad level in dir2 node d299df7... xfs_repair: avoid segfault if reporting progress early in repair e660c2e... Sent: Friday, October 11, 2013 12:38:11 AM Subject:
Bug#725971: xfsprogs: config.guess/config.sub out of date for arm64

Can you fix the subject line of this commit, please?

Cheers,

Dave.

Signature · 7 lines
Dave Chinner
[email protected]

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs
#16 lore raw

Re: [ANNOUNCE] xfsprogs for-3.1.12 branch has been updated to 40c65a7

Quoted 10 lines · Dave Chinner
Quoted 7 lines · Rich Johnston

Arkadiusz,

Thanks for compiling the list.

The for-3.1.12 Branch created from "a547280... 3.1.11 release"

If there are others you want committed, please let me know.

What testing have you done on the branch?

Just a quick test. I wanted to get it out for others who are interested in this branch to start testing.

Quoted 15 lines
Quoted 14 lines

40c65a7... xfs_metadump: Make -F (force) optional
ec3a1f1... xfs_repair: suggest reboot after dangerous repair 8660952... xfsprogs: suggest "-d" option for repair of RO mount 5a8206c... xfs_repair: correct docs for "-t" units
0d5c444... xfs_fsr: fix SWAPEXT failures under selinux f3d3dae... libxfs: fix root inode handling inconsistencies cf507b6... xfsprogs: remove xfs_check references from fsck.xfs script & manpage
6e2f2d1... xfsprogs: fix resource leak in longform_dir2_rebuild() ecde3f9... xfs_repair: test for bad level in dir2 node d299df7... xfs_repair: avoid segfault if reporting progress early in repair e660c2e... Sent: Friday, October 11, 2013 12:38:11 AM Subject:
Bug#725971: xfsprogs: config.guess/config.sub out of date for arm64

Quoted 1 line

Can you fix the subject line of this commit, please?

Sorry I missed that.

I can, but a git rebase --interactive e660c2e^ will change the commit ids after it. Is there another way to do this that does not change the commit ids?

Thanks
--Rich

Quoted 5 lines

Cheers,

Dave.

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs

#17 lore raw

Re: [ANNOUNCE] xfsprogs for-3.1.12 branch has been updated to 40c65a7

Hey Rich,

Quoted 30 lines · Rich Johnston
Quoted 10 lines · Dave Chinner
Quoted 7 lines · Rich Johnston

Arkadiusz,

Thanks for compiling the list.

The for-3.1.12 Branch created from "a547280... 3.1.11 release"

If there are others you want committed, please let me know.

What testing have you done on the branch?

Just a quick test. I wanted to get it out for others who are interested in this branch to start testing.

Quoted 15 lines
Quoted 14 lines

40c65a7... xfs_metadump: Make -F (force) optional
ec3a1f1... xfs_repair: suggest reboot after dangerous repair 8660952... xfsprogs: suggest "-d" option for repair of RO mount 5a8206c... xfs_repair: correct docs for "-t" units
0d5c444... xfs_fsr: fix SWAPEXT failures under selinux f3d3dae... libxfs: fix root inode handling inconsistencies cf507b6... xfsprogs: remove xfs_check references from fsck.xfs script & manpage
6e2f2d1... xfsprogs: fix resource leak in longform_dir2_rebuild() ecde3f9... xfs_repair: test for bad level in dir2 node d299df7... xfs_repair: avoid segfault if reporting progress early in repair e660c2e... Sent: Friday, October 11, 2013 12:38:11 AM Subject:
Bug#725971: xfsprogs: config.guess/config.sub out of date for arm64

Quoted 7 lines
Quoted 1 line

Can you fix the subject line of this commit, please?

Sorry I missed that.

I can, but a git rebase --interactive e660c2e^ will change the commit ids after it. Is there another way to do this that does not change the commit ids?

Any way you do it will change the commit ids. Just don't worry about it. If you really don't want to lose those old commits, you can create a tag for-3.1.12.bad to keep those old commits pinned in the repo, and then just force replace the branch after your rebase. Then you can remove the 'bad' tag after the release is completed. No big deal.

Regards,
Ben

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs

#18 lore raw

Re: [ANNOUNCE] xfsprogs for-3.1.12 branch has been updated to 40c65a7

In reply to Rich Johnston
Quoted 32 lines · Rich Johnston
Quoted 10 lines · Dave Chinner
Quoted 7 lines · Rich Johnston

Arkadiusz,

Thanks for compiling the list.

The for-3.1.12 Branch created from "a547280... 3.1.11 release"

If there are others you want committed, please let me know.

What testing have you done on the branch?

Just a quick test. I wanted to get it out for others who are interested in this branch to start testing.

Quoted 15 lines
Quoted 14 lines

40c65a7... xfs_metadump: Make -F (force) optional
ec3a1f1... xfs_repair: suggest reboot after dangerous repair 8660952... xfsprogs: suggest "-d" option for repair of RO mount 5a8206c... xfs_repair: correct docs for "-t" units
0d5c444... xfs_fsr: fix SWAPEXT failures under selinux f3d3dae... libxfs: fix root inode handling inconsistencies cf507b6... xfsprogs: remove xfs_check references from fsck.xfs script & manpage
6e2f2d1... xfsprogs: fix resource leak in longform_dir2_rebuild() ecde3f9... xfs_repair: test for bad level in dir2 node d299df7... xfs_repair: avoid segfault if reporting progress early in repair e660c2e... Sent: Friday, October 11, 2013 12:38:11 AM Subject:
Bug#725971: xfsprogs: config.guess/config.sub out of date for arm64

Quoted 7 lines
Quoted 1 line

Can you fix the subject line of this commit, please?

Sorry I missed that.

I can, but a git rebase --interactive e660c2e^ will change the commit ids after it. Is there another way to do this that does not change the commit ids?

Just rebase it - it's a test branch. In future, you might want to use a temporary branch for posting "work in progress" like this so that it's clear that it is a work in progress...

Cheers,

Dave.

Signature · 7 lines
Dave Chinner
[email protected]

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs
#19 lore raw

Re: [ANNOUNCE] xfsprogs for-3.1.12 branch has been updated to 40c65a7

In reply to Rich Johnston
Quoted 16 lines · Rich Johnston

Arkadiusz,

Thanks for compiling the list.

The for-3.1.12 Branch created from "a547280... 3.1.11 release"

If there are others you want committed, please let me know.

Thanks
--Rich

The following have been committed from the candidate list provided by Arkadiusz:

Quoted 6 lines

40c65a7... xfs_metadump: Make -F (force) optional ec3a1f1... xfs_repair: suggest reboot after dangerous repair 8660952... xfsprogs: suggest "-d" option for repair of RO mount 5a8206c... xfs_repair: correct docs for "-t" units 0d5c444... xfs_fsr: fix SWAPEXT failures under selinux f3d3dae... libxfs: fix root inode handling inconsistencies

Introduces this compiler warning in db/init.c:

init.c: In function ¿init¿:
init.c:152:3: warning: implicit declaration of function ¿xfs_initialize_perag_data¿ [-Wimplicit-function-declaration]

Quoted 6 lines

cf507b6... xfsprogs: remove xfs_check references from fsck.xfs script & manpage
6e2f2d1... xfsprogs: fix resource leak in longform_dir2_rebuild() ecde3f9... xfs_repair: test for bad level in dir2 node d299df7... xfs_repair: avoid segfault if reporting progress early in repair e660c2e... Sent: Friday, October 11, 2013 12:38:11 AM Subject:

Quoted 8 lines

Bug#725971: xfsprogs: config.guess/config.sub out of date for arm64 5e1c7ac... xfsprogs: handle symlinks etc in fs_table_initialise_mounts() aba6204... xfsprogs: fix return value of verify_set_primary_sb() a2ea195... xfsprogs: fix potential memory leak in verify_set_primary_sb() 3e53939... xfsprogs: fix potential memory leak in repare/sb.c 820b8c2... xfsprogs: avoid array overflow in pf_batch_read() 2a1dcdb... xfs_repair: zero out unused parts of superblocks 982508b... xfsprogs: fix inode crash in xfs_repair

Introduces this compiler warning in repair/phase6.c:

phase6.c: In function ¿mk_orphanage¿:
phase6.c:895:3: warning: implicit declaration of function ¿alloc_ex_data¿ [-Wimplicit-function-declaration]

Running some smoke testing on the branch now. Rich, I expect that you'll be doing some in-depth functionality, scalability and stress testing of the branch over the next week or two?

Cheers,

Dave.

Signature · 7 lines
Dave Chinner
[email protected]

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs
#20 lore raw

Re: [ANNOUNCE] xfsprogs for-3.1.12 branch has been updated to 40c65a7

Quoted 45 lines · Dave Chinner
Quoted 16 lines · Rich Johnston

Arkadiusz,

Thanks for compiling the list.

The for-3.1.12 Branch created from "a547280... 3.1.11 release"

If there are others you want committed, please let me know.

Thanks
--Rich

The following have been committed from the candidate list provided by Arkadiusz:

Quoted 6 lines

40c65a7... xfs_metadump: Make -F (force) optional ec3a1f1... xfs_repair: suggest reboot after dangerous repair 8660952... xfsprogs: suggest "-d" option for repair of RO mount 5a8206c... xfs_repair: correct docs for "-t" units 0d5c444... xfs_fsr: fix SWAPEXT failures under selinux f3d3dae... libxfs: fix root inode handling inconsistencies

Introduces this compiler warning in db/init.c:

init.c: In function ¿init¿:
init.c:152:3: warning: implicit declaration of function ¿xfs_initialize_perag_data¿ [-Wimplicit-function-declaration]

Quoted 6 lines

cf507b6... xfsprogs: remove xfs_check references from fsck.xfs script & manpage
6e2f2d1... xfsprogs: fix resource leak in longform_dir2_rebuild() ecde3f9... xfs_repair: test for bad level in dir2 node d299df7... xfs_repair: avoid segfault if reporting progress early in repair e660c2e... Sent: Friday, October 11, 2013 12:38:11 AM Subject:

Quoted 8 lines

Bug#725971: xfsprogs: config.guess/config.sub out of date for arm64 5e1c7ac... xfsprogs: handle symlinks etc in fs_table_initialise_mounts() aba6204... xfsprogs: fix return value of verify_set_primary_sb() a2ea195... xfsprogs: fix potential memory leak in verify_set_primary_sb() 3e53939... xfsprogs: fix potential memory leak in repare/sb.c 820b8c2... xfsprogs: avoid array overflow in pf_batch_read() 2a1dcdb... xfs_repair: zero out unused parts of superblocks 982508b... xfsprogs: fix inode crash in xfs_repair

Quoted 8 lines

Introduces this compiler warning in repair/phase6.c:

phase6.c: In function ¿mk_orphanage¿:
phase6.c:895:3: warning: implicit declaration of function ¿alloc_ex_data¿ [-Wimplicit-function-declaration]

Running some smoke testing on the branch now. Rich, I expect that you'll be doing some in-depth functionality, scalability and stress testing of the branch over the next week or two?

I will do *some* testing yes. This is more for those who are interested in this branch. I do not plan on using it myself. I put it out there for those interested in a 3.1.12 release.

Thanks
--Rich

Quoted 5 lines

Cheers,

Dave.

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs

#21 lore raw

Re: [ANNOUNCE] xfsprogs for-3.1.12 branch has been updated to 40c65a7

Quoted 51 lines · Rich Johnston
Quoted 50 lines · Dave Chinner
Quoted 16 lines · Rich Johnston

Arkadiusz,

Thanks for compiling the list.

The for-3.1.12 Branch created from "a547280... 3.1.11 release"

If there are others you want committed, please let me know.

Thanks
--Rich

The following have been committed from the candidate list provided by Arkadiusz:

Quoted 6 lines

40c65a7... xfs_metadump: Make -F (force) optional ec3a1f1... xfs_repair: suggest reboot after dangerous repair 8660952... xfsprogs: suggest "-d" option for repair of RO mount 5a8206c... xfs_repair: correct docs for "-t" units 0d5c444... xfs_fsr: fix SWAPEXT failures under selinux f3d3dae... libxfs: fix root inode handling inconsistencies

Introduces this compiler warning in db/init.c:

init.c: In function ¿init¿:
init.c:152:3: warning: implicit declaration of function ¿xfs_initialize_perag_data¿ [-Wimplicit-function-declaration]

Quoted 6 lines

cf507b6... xfsprogs: remove xfs_check references from fsck.xfs script & manpage
6e2f2d1... xfsprogs: fix resource leak in longform_dir2_rebuild() ecde3f9... xfs_repair: test for bad level in dir2 node d299df7... xfs_repair: avoid segfault if reporting progress early in repair e660c2e... Sent: Friday, October 11, 2013 12:38:11 AM Subject:

Quoted 8 lines

Bug#725971: xfsprogs: config.guess/config.sub out of date for arm64 5e1c7ac... xfsprogs: handle symlinks etc in fs_table_initialise_mounts() aba6204... xfsprogs: fix return value of verify_set_primary_sb() a2ea195... xfsprogs: fix potential memory leak in verify_set_primary_sb() 3e53939... xfsprogs: fix potential memory leak in repare/sb.c 820b8c2... xfsprogs: avoid array overflow in pf_batch_read() 2a1dcdb... xfs_repair: zero out unused parts of superblocks 982508b... xfsprogs: fix inode crash in xfs_repair

Introduces this compiler warning in repair/phase6.c:

phase6.c: In function ¿mk_orphanage¿:
phase6.c:895:3: warning: implicit declaration of function ¿alloc_ex_data¿ [-Wimplicit-function-declaration]

Quoted 7 lines
Quoted 3 lines

Running some smoke testing on the branch now. Rich, I expect that you'll be doing some in-depth functionality, scalability and stress testing of the branch over the next week or two?

I will do *some* testing yes. This is more for those who are interested in this branch. I do not plan on using it myself. I put it out there for those interested in a 3.1.12 release.

You're acting as the maintainer of this branch, so it's primarily your responsibility to make sure it works. That means you need to ensure that it is fully tested and that when bugs and problems are supported you respond to them, triage them and fix them. It doesn't matter if you intend to use the branch or not - you signed up to do this for the XFS users that were asking for it and that means you need to ensure the quality of the release is acceptible. Being a maintainer is more than being allowed to commit patches to a repository.

Cheers,

Dave.

Signature · 7 lines
Dave Chinner
[email protected]

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs
author
#22 lore raw

Re: [ANNOUNCE] xfsprogs for-3.1.12 branch has been updated to 40c65a7

In reply to Dave Chinner
Quoted 7 lines · Dave Chinner
Quoted 1 line

f3d3dae... libxfs: fix root inode handling inconsistencies

Introduces this compiler warning in db/init.c:

init.c: In function ¿init¿:
init.c:152:3: warning: implicit declaration of function ¿xfs_initialize_perag_data¿ [-Wimplicit-function-declaration]

Probably revert this one, right?

Quoted 7 lines
Quoted 1 line

982508b... xfsprogs: fix inode crash in xfs_repair

Introduces this compiler warning in repair/phase6.c:

phase6.c: In function ¿mk_orphanage¿:
phase6.c:895:3: warning: implicit declaration of function ¿alloc_ex_data¿ [-Wimplicit-function-declaration

2e1622703bb5c01e5f6d5fa087999c768868aa0b xfs_repair: add prototype for alloc_ex_data()

Signature · 6 lines
Arkadiusz Miśkiewicz, arekm / maven.pl

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs
#23 lore raw

Re: [ANNOUNCE] xfsprogs for-3.1.12 branch has been updated to 40c65a7

Hey Arkadiusz,

Quoted 11 lines · Arkadiusz Miśkiewicz
Quoted 7 lines · Dave Chinner
Quoted 1 line

f3d3dae... libxfs: fix root inode handling inconsistencies

Introduces this compiler warning in db/init.c:

init.c: In function ¿init¿:
init.c:152:3: warning: implicit declaration of function ¿xfs_initialize_perag_data¿ [-Wimplicit-function-declaration]

Probably revert this one, right?

Just a compiler warning which does not happen in the master branch.

Quoted 11 lines
Quoted 8 lines
Quoted 2 lines

982508b... xfsprogs: fix inode crash in xfs_repair

Introduces this compiler warning in repair/phase6.c:

phase6.c: In function ¿mk_orphanage¿:
phase6.c:895:3: warning: implicit declaration of function ¿alloc_ex_data¿ [-Wimplicit-function-declaration

2e1622703bb5c01e5f6d5fa087999c768868aa0b xfs_repair: add prototype for alloc_ex_data()

This has been committed

e3ffe68... xfs_repair: add prototype for alloc_ex_data()

Thanks
--Rich

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs

#24 lore raw

Re: [ANNOUNCE] xfsprogs for-3.1.12 branch has been updated to 40c65a7

Quoted 10 lines · Rich Johnston

Hey Arkadiusz,

Quoted 6 lines · Arkadiusz Miśkiewicz
Quoted 4 lines · Dave Chinner
Quoted 1 line

f3d3dae... libxfs: fix root inode handling inconsistencies

Introduces this compiler warning in db/init.c:

Quoted 7 lines
Quoted 5 lines
Quoted 3 lines

init.c: In function ¿init¿:
init.c:152:3: warning: implicit declaration of function ¿xfs_initialize_perag_data¿ [-Wimplicit-function-declaration]

Probably revert this one, right?

Just a compiler warning which does not happen in the master branch.

well, let's get it fixed anyway.

moving the prototype to include/xfs_mount.h fixes it up.

-Eric

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs

#25 lore raw

Re: [ANNOUNCE] xfsprogs for-3.1.12 branch has been updated to 40c65a7

Quoted 17 lines · Eric Sandeen
Quoted 16 lines · Rich Johnston

Hey Arkadiusz,

Quoted 11 lines · Arkadiusz Miśkiewicz
Quoted 7 lines · Dave Chinner
Quoted 1 line

f3d3dae... libxfs: fix root inode handling inconsistencies

Introduces this compiler warning in db/init.c:

init.c: In function ¿init¿:
init.c:152:3: warning: implicit declaration of function ¿xfs_initialize_perag_data¿ [-Wimplicit-function-declaration]

Probably revert this one, right?

Quoted 8 lines
Quoted 1 line

Just a compiler warning which does not happen in the master branch.

well, let's get it fixed anyway.

moving the prototype to include/xfs_mount.h fixes it up.

-Eric

No problem added to the cherry picked commit.

New commit id.
c21f3b5... libxfs: fix root inode handling inconsistencies

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs

#26 lore raw

Re: [ANNOUNCE] xfsprogs for-3.1.12 branch has been updated to 40c65a7

In reply to Rich Johnston

Hey Folks,

Has anyone been testing this branch, any feedback? There are still a couple of platforms I would like to test before releasing this branch.

Unless there are objections or problems I plan to have all my testing done by Wednesday February 26.

Thanks
--Rich

Quoted 53 lines · Rich Johnston

Arkadiusz,

Thanks for compiling the list.

The for-3.1.12 Branch created from "a547280... 3.1.11 release"

If there are others you want committed, please let me know.

Thanks
--Rich

The following have been committed from the candidate list provided by Arkadiusz:

40c65a7... xfs_metadump: Make -F (force) optional
ec3a1f1... xfs_repair: suggest reboot after dangerous repair 8660952... xfsprogs: suggest "-d" option for repair of RO mount 5a8206c... xfs_repair: correct docs for "-t" units
0d5c444... xfs_fsr: fix SWAPEXT failures under selinux f3d3dae... libxfs: fix root inode handling inconsistencies cf507b6... xfsprogs: remove xfs_check references from fsck.xfs script & manpage
6e2f2d1... xfsprogs: fix resource leak in longform_dir2_rebuild() ecde3f9... xfs_repair: test for bad level in dir2 node d299df7... xfs_repair: avoid segfault if reporting progress early in repair e660c2e... Sent: Friday, October 11, 2013 12:38:11 AM Subject:
Bug#725971: xfsprogs: config.guess/config.sub out of date for arm64 5e1c7ac... xfsprogs: handle symlinks etc in fs_table_initialise_mounts() aba6204... xfsprogs: fix return value of verify_set_primary_sb() a2ea195... xfsprogs: fix potential memory leak in verify_set_primary_sb() 3e53939... xfsprogs: fix potential memory leak in repare/sb.c 820b8c2... xfsprogs: avoid array overflow in pf_batch_read() 2a1dcdb... xfs_repair: zero out unused parts of superblocks 982508b... xfsprogs: fix inode crash in xfs_repair
b8890d7... xfsprogs: fix unint var in repair phase6
7d710b9... xfsprogs: fix agcnts leak in xfs_repair's scan_ags 291c159... xfsprogs:free bp in xlog_find_tail() error path 40b9d95... xfsprogs: free bp in xlog_find_zeroed() error path a9b4757... xfsprogs: fix buffer leak in xlog_print_find_oldest 32aae5b... xfsprogs: avoid double-free in xfs_attr_node_addname 243cf82... mkfs.xfs: fix protofile name create block reservation b31a23d... xfs_metadump: manpage fix regarding frozen fs f8b80a3... xfs_fsr: file reads should be O_DIRECT

The following commits were not applicable.
e7c05095, e6efb967 6aa3d87b

Quoted 4 lines

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs

#27 lore raw

Re: [ANNOUNCE] xfsprogs for-3.1.12 branch has been updated to 40c65a7

Quoted 5 lines · Rich Johnston

Hey Folks,

Has anyone been testing this branch, any feedback? There are still a couple of platforms I would like to test before releasing this branch.

No, I haven't looked at it for weeks. Too busy working on 3.2 and other stuff at the moment.

Quoted 2 lines

Unless there are objections or problems I plan to have all my testing done by Wednesday February 26.

Can you outline the testing you have done?

Cheers,

Dave.

Signature · 7 lines
Dave Chinner
[email protected]

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs
#28 lore raw

Re: [ANNOUNCE] xfsprogs for-3.1.12 branch has been updated to 40c65a7

On 02/19/2014 03:50 PM, Dave Chinner wrote:
...

Quoted 2 lines

Can you outline the testing you have done?

Sorry it took so long to get back.

I used Kernel version 3.14.0-rc4+
Compared the results against TOT xfsprogs and for-3.1.12

Most of the bugs were log and repair specific so I ran a bunch of tests to exercise those.

I ran xfstests several times on ia64, x86_64 and i386 with the following command:
./check -g auto -x dmapi -x freeze -x tape

generic/306 - output mismatch (see /usr/src/3.1.12/xfstests/results//generic/306.out.bad)

    --- tests/generic/306.out    2014-03-11 09:42:47.000000000 -0500
    +++ /usr/src/3.1.12/xfstests/results//generic/306.out.bad 
2014-03-11 11:34:01.000000000 -0500
    @@ -2,11 +2,9 @@
     == try to create new file
     touch: cannot touch 'SCRATCH_MNT/this_should_fail': Read-only file

system
== pwrite to null device
-wrote 512/512 bytes at offset 0
-XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)xfs/001 [not run] mkfs.xfs doesn't have crc feature

xfs/002     [not run] mkfs.xfs doesn't have crc feature
xfs/005     [not run] mkfs.xfs doesn't have crc feature
xfs/299     [not run] mkfs.xfs doesn't have crc feature
xfs/304     [not run] mkfs.xfs doesn't have crc feature
xfs/305     [not run] mkfs.xfs doesn't have crc feature
    +XFS_IOC_FSGEOMETRY: Inappropriate ioctl for device
     == pread from zero device
    ...
    (Run 'diff -u tests/generic/306.out

/usr/src/3.1.12/xfstests/results//generic/306.out.bad' to see the entire diff)

xfs/016 - output mismatch (see /usr/src/3.1.12/xfstests/resuxfs/lts//xfs/016.out.bad)

    --- tests/xfs/016.out    2014-03-11 09:42:47.000000000 -0500
    +++ /usr/src/3.1.12/xfstests/results//xfs/016.out.bad    2014-03-11 
11:34:15.000000000 -0500
    @@ -11,25 +11,30 @@
        *** mount
        *** fiddle
        *** unmount
    +XFS: Invalid block length (0x2000) given for buffer
     *** generate log traffic
        *** mount
        *** fiddle
    ...
    (Run 'diff -u tests/xfs/016.out

/usr/src/3.1.12/xfstests/results//xfs/016.out.bad' to see the entire diff)

xfs/122 - output mismatch (see /usr/src/3.1.12/xfstests/results//xfs/122.out.bad)

    --- tests/xfs/122.out    2014-03-11 09:42:47.000000000 -0500
    +++ /usr/src/3.1.12/xfstests/results//xfs/122.out.bad    2014-03-11

11:34:17.000000000 -0500
@@ -49,39 +49,70 @@
sizeof( xfs_agfl_t ) = 4
sizeof( xfs_agi_t ) = 296
sizeof( xfs_alloc_rec_incore_t ) = 8
+sizeof( xfs_alloc_rec_incore_t ) = 8
+sizeof( xfs_alloc_rec_t ) = 8
sizeof( xfs_alloc_rec_t ) = 8
sizeof( xfs_attr_inactive_list_t ) = 8
...
(Run 'diff -u tests/xfs/122.out
/usr/src/3.1.12/xfstests/results//xfs/122.out.bad' to see the entire diff)

- output mismatch (see
/usr/src/3.1.12/xfstests/results//xfs/178.out.bad)

     --- tests/xfs/178.out    2014-03-11 09:42:47.000000000 -0500
     +++ /usr/src/3.1.12/xfstests/results//xfs/178.out.bad    2014-03-11

12:09:55.000000000 -0500
@@ -10,6 +10,9 @@

attempting to find secondary superblock...
found candidate secondary superblock...
+error reading superblock 1 -- seek to offset 1099511623680 failed +unable to verify superblock, continuing...
+found candidate secondary superblock...
verified secondary superblock...

Ran: generic/001 generic/002 generic/003 generic/005 generic/006 generic/007 generic/010 generic/011 generic/013 generic/014 generic/015 generic/020 generic/053 generic/062 generic/069 generic/070 generic/074 generic/075 generic/076 generic/077 generic/079 generic/083 generic/088 generic/089 generic/091 generic/100 generic/105 generic/112 generic/113 generic/117 generic/120 generic/123 generic/124 generic/125 generic/126 generic/127 generic/128 generic/129 generic/130 generic/131 generic/132 generic/133 generic/135 generic/141 generic/169 generic/184 generic/192 generic/193 generic/198 generic/204 generic/207 generic/208 generic/209 generic/210 generic/211 generic/212 generic/213 generic/214 generic/215 generic/219 generic/221 generic/223 generic/224 generic/225 generic/226 generic/228 generic/230 generic/231 generic/232 generic/233 generic/234 generic/235 generic/236 generic/237 generic/239 generic/240 generic/245 generic/246 generic/247 generic/248 generic/249 generic/255 generic/256 generic/257 generic/258 generic/263 generic/269 generic/270 generic/273 generic/274 generic/275 generic/277 generic/285 generic/286 generic/294 generic/299 generic/300 generic/306 generic/307 generic/308 generic/309 generic/310 generic/312 generic/313 generic/314 generic/315 generic/316 generic/317 generic/318 generic/319 generic/320 shared/032 shared/051 shared/218 shared/243 shared/298 xfs/003 xfs/004 xfs/008 xfs/009 xfs/012 xfs/016 xfs/017 xfs/019 xfs/021 xfs/026 xfs/027 xfs/028 xfs/029 xfs/030 xfs/031 xfs/033 xfs/034 xfs/041 xfs/042 xfs/045 xfs/046 xfs/047 xfs/048 xfs/049 xfs/050 xfs/052 xfs/054 xfs/056 xfs/059 xfs/060 xfs/061 xfs/063 xfs/064 xfs/065 xfs/066 xfs/067 xfs/071 xfs/072 xfs/073 xfs/078 xfs/084 xfs/085 xfs/086 xfs/087 xfs/092 xfs/096 xfs/103 xfs/104 xfs/108 xfs/109 xfs/110 xfs/116 xfs/118 xfs/121 xfs/122 xfs/134 xfs/137 xfs/138 xfs/139 xfs/140 xfs/164 xfs/165 xfs/166 xfs/167 xfs/170 xfs/174 xfs/178 xfs/179 xfs/180 xfs/181 xfs/182 xfs/183 xfs/186 xfs/187 xfs/188 xfs/189 xfs/190 xfs/194 xfs/195 xfs/196 xfs/199 xfs/200 xfs/201 xfs/202 xfs/203 xfs/205 xfs/206 xfs/216 xfs/217 xfs/220 xfs/222 xfs/227 xfs/229 xfs/238 xfs/242 xfs/244 xfs/250 xfs/252 xfs/253 xfs/259 xfs/261 xfs/262 xfs/266 xfs/278 xfs/281 xfs/282 xfs/283 xfs/287 xfs/290 xfs/291 xfs/292 xfs/293 xfs/295 xfs/296 xfs/298 xfs/301 xfs/302 xfs/303 xfs/306 Not run: generic/093 generic/097 generic/099 generic/241 generic/260 generic/288 generic/311 generic/321 generic/322 shared/001 shared/002 shared/003 shared/004 shared/005 shared/272 shared/289 xfs/001 xfs/002 xfs/005 xfs/040 xfs/044 xfs/057 xfs/058 xfs/090 xfs/094 xfs/095 xfs/148 xfs/149 xfs/177 xfs/191 xfs/197 xfs/279 xfs/299 xfs/300 xfs/304 xfs/305
Failures: generic/306 xfs/016 xfs/122 xfs/178
Failed 4 of 235 tests

There were some tests that did not run that were crc specific

xfs/001     [not run] mkfs.xfs doesn't have crc feature
xfs/002     [not run] mkfs.xfs doesn't have crc feature
xfs/005     [not run] mkfs.xfs doesn't have crc feature
xfs/299     [not run] mkfs.xfs doesn't have crc feature
xfs/304     [not run] mkfs.xfs doesn't have crc feature
xfs/305     [not run] mkfs.xfs doesn't have crc feature

generic/306
There is on patch that could be cherry-picked to fix it. (I re-ran tests with it applied.

commit fcd6fa7f6a29188a756ff8b071a152bc9d62baf4

Author: Eric Sandeen <[email protected]>
Date:   Tue Oct 8 15:17:50 2013 +0000

xfsprogs: restrict platform_test_xfs_fd to regular files

xfs/016 passes
If I revert bb4074a31 xfstests: use minimum log size in xfs/016
http://oss.sgi.com/archives/xfs/2013-10/msg00559.html

xfs/122 also fails on both and most people don't install INDENT so they would never see it.

xfs/178
This test blows away the main sb, then does repair. It should have found sb1 but failed, it did find sb2. strange but not an error. I got the same error using TOT xfsprogs.

Thanks
--Rich

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs

#29 lore raw
In reply to Arkadiusz Miśkiewicz
Quoted 8 lines · Arkadiusz Miśkiewicz

Hello.

What's needed for both 3.1.12 and 3.2.0 releases to happen?

If 3.2.0 is to be released soon then nice. Dave, AFAIK it is waiting for you and xfs_db crc work only?

Quoted 7 lines

If 3.2.0 is far away then it would be nice to see 3.1.12 with bugfixes only (for those paranoid about new stuff) but 3.1.13 very soon after that with backported features like project quota support, parallelism, performance improvements.

Obviously I hope 3.2.0 is soon and there is no need to waste effort on 3.1.x.

Is there any interest in a 3.1.12 release or will a 3.2.x release happen soon?

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs

#30 lore raw
Quoted 14 lines · Rich Johnston
Quoted 13 lines · Arkadiusz Miśkiewicz

Hello.

What's needed for both 3.1.12 and 3.2.0 releases to happen?

If 3.2.0 is to be released soon then nice. Dave, AFAIK it is waiting for you and xfs_db crc work only?

If 3.2.0 is far away then it would be nice to see 3.1.12 with bugfixes only (for those paranoid about new stuff) but 3.1.13 very soon after that with backported features like project quota support, parallelism, performance improvements.

Quoted 4 lines
Quoted 2 lines

Obviously I hope 3.2.0 is soon and there is no need to waste effort on 3.1.x.

Is there any interest in a 3.1.12 release or will a 3.2.x release happen soon?

3.2.x should happen soon. I've got a few patches to integrate and test and then we can move to a 3.2-rc1 release and go from there.

Cheers,

Dave.

Signature · 7 lines
Dave Chinner
[email protected]

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs
#31 lore raw
Quoted 15 lines · Dave Chinner
Quoted 14 lines · Rich Johnston
Quoted 13 lines · Arkadiusz Miśkiewicz

Hello.

What's needed for both 3.1.12 and 3.2.0 releases to happen?

If 3.2.0 is to be released soon then nice. Dave, AFAIK it is waiting for you and xfs_db crc work only?

If 3.2.0 is far away then it would be nice to see 3.1.12 with bugfixes only (for those paranoid about new stuff) but 3.1.13 very soon after that with backported features like project quota support, parallelism, performance improvements.

Quoted 7 lines
Quoted 4 lines
Quoted 2 lines

Obviously I hope 3.2.0 is soon and there is no need to waste effort on 3.1.x.

Is there any interest in a 3.1.12 release or will a 3.2.x release happen soon?

3.2.x should happen soon. I've got a few patches to integrate and test and then we can move to a 3.2-rc1 release and go from there.

Awesome, thanks Dave.

Quoted 5 lines

Cheers,

Dave.

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs