Activity
From 2012-07-23 to 2012-08-21
2012-08-21
-
06:35 PM Feature #3104 (Resolved): eliminate empty bpobjs
- From Matt Ahrens's original bug report at delphix:
Every snapshot has a deadlist, which tracks which blocks it del... -
12:45 PM Bug #2986: nfs: exi refcounter leak at rfs3_lookup
- ...
-
11:43 AM Bug #3086 (Resolved): unnecessarily setting DS_FLAG_INCONSISTENT on async destroyed datasets
- changeset: 13776:cd512c80fd75
tag: tip
user: Matthew Ahrens <mahrens@delphix.com>
date: Tu... -
- The refcount structure is different when compiled debug vs nondebug. Therefore, we want to make sure we get the refc...
2012-08-20
-
11:56 PM Feature #3057 (Resolved): Remove sgml util from cmd/man
- Resolved in r13775 commit:1d7e7613366b
-
- While running 'ztest' we occasionally hit failures that look like this:
Executing zdb -bccsv -U /tank/fs/ztest/zpo... -
07:27 PM Feature #104: Bring back lx brand
- As recently pointed on the mailing list, there is a blog with diff-files and explanation of these updates which retur...
-
- When renaming a zvol we need to sync out dirty data first.
-
02:29 AM Bug #3099 (Closed): AMD hudson D3 AHCI support
- 151a5 not support AMD hudson D3 chipset AHCI.
AMD hudson D3 chipset AHCI device, the vendor id is 0x1022, device i... -
- ...
2012-08-19
-
10:05 PM Bug #3097 (New): zfs set doesn't check length of user property
- Despite the zfs(1M) saying the following in 'User Properties' section:...
-
- perl and tcpd. I'll have the webrev out tomorrow.
-
07:33 PM Bug #3088: some THIRDPARTYLICENSE files erroneously removed
- It would be nice to have the list of them here ;-)
-
03:35 PM Bug #3096 (Resolved): libctf does not validate function types correctly
- Currently libctf does not validate the types of a function that are passed in. It does not make sure that the return ...
-
- Currently libctf does no validation when using the ctf_create_array() function. All it currently checks is that the a...
-
- As part of doing the work for 3089 and trying to support self-referential structures, this forces us to be unable to ...
-
- zfs {user,group}space's -i option (SID to POSIX ID translation) isn't used at all.
2012-08-18
-
02:49 PM Bug #3091 (Resolved): add -n to zlogin so its more compatible with rsh command line
- In some circumstances, it is easy to replace the execution of rsh with zlogin quite easily, e.g.
rsh remotemachine... -
- Resolved in r13774 commit:80b84226b49f
2012-08-17
-
- There is a period between the time spa_change_guid() is called and when the
labels with the new guid are synced to d... -
- Often times when using mdb, people run into situations where they either don't have CTF data or there are anonymous s...
-
- A couple of THIRDPARTYLICENSE files got removed in error, despite the fact we still use them.
"Luckily", the close... -
03:56 PM Bug #3087: libuuid has a lot of dependencies
- Rich Lowe wrote:
> libuuid is not the problem, libdlpi (and others) are, as you point out. Fix them, if anything.
... -
- libuuid is not the problem, libdlpi (and others) are, as you point out. Fix them, if anything.
-
- libuuid has a lot of dependencies through libdlpi.so:...
2012-08-16
-
- The DS_FLAG_INCONSISTENT is set synchronously on a dataset before it is
destroyed. With async_destroy enabled this i... -
- Resolved in r13773 commit:00c2a08cf1bb
-
-
- Resolved in r13772 commit:2579580ac955
-
- It looks like the panic will occur any time a 'zfs release' causes a defer-destroyed snapshot to actually get destroy...
-
- I booted a prior BE, out of desparation, which hasn't crashed.
It seems likely that Matt's change to use _internal... -
- Yes, dsl_dataset_user_release_sync....
-
- I think we found the problem. spa_history_log_internal_ds() is being called too late from dsl_dataset_user_release_sy...
-
01:32 AM Bug #3085: zfs diff panics, then panics in a loop on booting
- Are you able to get a dump? If the diff was not on your root filesystem, you can boot with "-m milestone=none" in th...
-
- It's not clear to me (or, I think, Schrock), whether the problem is that the ds_dir is NULL, or that we don't cope.
... -
- It is perhaps important that the zfs diff invocation I used was of the form...
-
- I had the misfortune to interrupt a 'zfs diff'. The system immediately panicked for a reason I didn't catch, subsequ...
-
01:55 PM Feature #212 (Resolved): Atheros AR8132 / L1c Gigabit Ethernet Adapter
- Integrated.
changeset: 13768:ed21ea5d20cf
tag: tip
user: Gary Mills <gary_mills@fastmail.fm>
d... -
10:13 AM Bug #3084 (Feedback): System slow when accssing many file using CIFS.
- Here is some fresh data from this morning:...
2012-08-15
-
- h2. General Description
The entire system becomes 'slow' during(possible also after) accessing many files in a large... -
-
- Resolved in r13771 commit:2903afd49fe8
-
02:56 PM Bug #2601: kctl_isadep.c should support four serial consoles
- The change to kctl_pcache_create() is pretty simple.
I've tested this change on a system with a Tyan motherboard t...
2012-08-13
-
- cleanup in dcopy_free is uninitialized, if the stack garbage under it is non-0, we'll free dcopy channels which we sh...
-
2012-08-12
-
- changeset: 13770:f6f7025c7a4d
tag: tip
user: Igor Kozhukhov <ikozhukhov@gmail.com>
date: S... -
12:59 PM Bug #3068 (In Progress): vmstat swap/free wrong in local zones
2012-08-11
-
- Is this not just the problem of the root password being expired initially? In that case, changing the root password ...
-
04:14 PM Bug #3072 (New): GUI of TimeSlider will not start until password change
- As reported by openindiana@redharvey.com :
boot from openindiana 151a5 desktop dvd
default keyboard, language
la...
2012-08-10
-
08:10 PM Bug #3070 (In Progress): write lgrpinfo by C based on Perl lgrpinfo.pl
-
07:10 PM Bug #3013 (Resolved): ifconfig with no args should list all interfaces
- Fixed in:
changeset: 13769:2317a7492de4
tag: tip
user: Daniil Lunev <d.lunev.mail@gmail.com>
dat... -
12:37 PM Bug #3052: uname compatibility should be implemented with redefine_extname
- [spam was here]
-
- Resolved in r13767 commit:8c906b14afbd
-
09:18 AM Bug #3055: sendfile sets *off to 0 when interrupted by SIGALRM
- Daniil Lunev wrote:
> Server of it. Can you test these server and client? Will they reproduce the bug?
I've nearl... -
09:06 AM Bug #3055: sendfile sets *off to 0 when interrupted by SIGALRM
- Server of it. Can you test these server and client? Will they reproduce the bug?...
2012-08-09
-
- vmstat sometimes displays incorrect values when used inside a local zone with 1 second interval and count set:
<pr... -
- May be, I am doing something wrong, but off don't become zero in the followed code.
@while true; do kill -SIGALRM PI... -
- Daniil Lunev wrote:
> I call sendfile with non-blocking socket, it catch SIGALARM. Offset isn't changed. What versio...
2012-08-08
-
- I call sendfile with non-blocking socket, it catch SIGALARM. Offset isn't changed. What version of OI do you use?
-
- Daniil Lunev wrote:
> I can't reproduce this bug on my system. It works pretty fine. I call sendfilev64, then send S... -
- And sendfile64, even after receiving SIGALRM, returns right value - off + len;
-
- I can't reproduce this bug on my system. It works pretty fine. I call sendfilev64, then send SIGALRM and sfv_off isn'...
-
- Daniil Lunev wrote:
> libsendfile calls SYS_sendfile. This syscall check receiving any signals in three places. When... -
- libsendfile calls SYS_sendfile. This syscall check receiving any signals in three places. When it finds out that ther...
-
- I think the justification here is that the code was used for testing a specific scenario during implementation and is...
-
03:19 PM Bug #3067 (Resolved): Typo in spelling "succssful"
- There are another two occurrences of misspelled "succssful" word beyond the one mentioned in 3064...
-
11:42 AM Support #3066 (New): ahci: [ID 517647 kern.warning] WARNING: ahci0: watchdog port 1 satapkt 0xffffff02ea521988 timed out
- not sure if this is a bug, a regression in oi_151a5 or my ssd too close to full or worse, slowly dying.
I notice a... -
09:02 AM Bug #3065 (Resolved): some functions in the tcp module can be static
- Some functions declared as extern are in fact local static ones.
-
- Line 3274 misspells the word "successful"...
-
- Rich Lowe wrote:
> Resolved in r13761 commit:229af35d14fb
>
> Unless there's some reason this was left open? Dan?... -
- Resolved in r13761 commit:229af35d14fb
Unless there's some reason this was left open? Dan?
2012-08-07
-
-
- I can't change state "to RTI"
-
- Throughout the tree, there are quite a few places where we compare an enum foo to an enum bar, almost always accident...
-
- Given the long lead time necessary to get everything fixed and tested, I've begun trying to get things happy with 4.6...
-
- Daniil Lunev wrote:
> And what about other signals? Do they reset the offset?
Don't know, SIGALRM is the only sig... -
- And what about other signals? Do they reset the offset?
-
- Igor Pashev wrote:
> Yuri Pankov wrote:
> > there's a column named *USED* in the output, and it should report corre...
2012-08-06
-
- zfs {user,group}space -o doesn't use specified fields order (output is still in the default order)....
-
-
- For original kernel, dtrace-1000, stack:
unix`default_lock_delay+0x8c
unix`mutex_vec... -
- can I send it to RTI ?
-
- can I send it to RTI ?
-
- They are files have no correct prototypes of functions:
cmd/tcpd/fakelog.c
have:
openlog(name, logopt, facility)
...
2012-08-05
-
- Anything stopping this from going forward (as provided workaround doesn't really work)?
-
- Just remove usr/src/cmd/man/src/util. It should be strictly NOOP as current man(1) detects the presence of sgml2roff ...
2012-08-04
-
04:49 PM Feature #3056 (New): Support for Broadcom 5720
-
- the same problem with #3053
I have tested builds of tools by 64bit compiler.
I have made wrapper for it with tr... -
- tools/protolist/protolist.c should use __x86 for intel platforms because have problems with builds by 64bit compiler.
-
- I have tested builds of tools by 64bit compiler.
I have made wrapper for it with translate build line:
/usr/gcc/4... -
- files at usr/src/tools/protocmp/* should use __x86 for intel platform.
we have problems with __i386 and i386 defines... -
- After spending several hours investigating a strange problem in Netatalk running on Openindiana I was finally able to...
-
- Here's what the fixed cpio does:...
2012-08-03
-
- At present, utsname.h contains two static function implementations of uname() and _uname() used to deal with binary c...
-
01:56 PM Feature #3047 (In Progress): grep support for -r would be useful
-
04:20 AM Bug #2560 (Resolved): ACPI parser fails to identify COM3 on Tyan motherboard
- ...
2012-08-02
-
- It should be (is) already fixed in https://www.xvoid.org/illumos/webrev/zfs-ugspace/.
-
- I suppose you could append a fake property that's never duplicated, and then omit it on the display. That should fix...
-
- The cause of the bug is using of avl tree in zfs(1M) to sort records. There are two ways to fix it:
1) update avl tr... -
- zfs {user,group}space [-s|-S] is broken (misses the entries with the same quota)....
-
- ...
-
- The /usr/bin/grep still does not support recursive string searches.
2012-08-01
-
- Resolved in r13765 commit:9410cf539b11
-
- This was caused by #3006, which I'll have backed out within an hour or so....
-
- The 'xzp->z_id' here should be 'zp->z_id', the xattr node may not exist, and has already been dealt with regardless:
... -
- I'm seeing this too, and will take a look.
-
07:25 PM Bug #3046 (Resolved): crash on boot in zfs_rmnode
- My system dumps core during boot with this changeset:
38b4aca480b3 tip
I took three pictures of my screen when ... -
- Backed out in 13765:9410cf539b11
2012-07-31
-
- Matthew Ahrens wrote:
> I saw ~ 1 million syscl/sec/CPU with 4 threads and 4 CPUs. Performance degraded with 8 thre... -
- I saw ~ 1 million syscl/sec/CPU with 4 threads and 4 CPUs. Performance degraded with 8 threads and 8 CPUs, to ~400K ...
-
- Sorry test it should be run in background
for i in `seq 1 16`; do
./test_stat file-$i &
done
-
- Matthew Ahrens wrote:
> I don't understand how you are pegging 16 CPUs with only 1 thread.
Matthew, I described i... -
- I don't understand how you are pegging 16 CPUs with only 1 thread.
I get 1.6 Million syscl/sec using your test (1 ... -
- b) rrwlock was introduced by commit:
changeset: 5326:6752aa2bd5bc
user: ek110237
date: ... -
- a) Accordingly to lockstat and dtrace outputs, a lot of time is spending at rrw_enter_read or rrw_exit, whereas
... -
- I have tested it with OI 151a5. Also, I have looked through the latest illumos-gate sources and it seems to show dele...
-
03:24 PM Bug #3043: Relax restriction on zfs_ioc_next_obj() iteration
- Upgraded to 151a5 and the problem is fixed.
Sorry for reporting an already fixed bug. -
- I'm working on OpenIndiana 151a4 and no deleted files or directories are showing up in zfs diff.
I will upgrade... -
- root@Exile:# zfs diff rpool/test@m1
M /rpool/test/mk
- /rpool/test/mk/testf
It seems to work as y... -
- changeset: 13764:38b4aca480b3
tag: tip
user: Madhav Suresh <madhav.suresh@delphix.com>
date: ... -
01:11 PM Bug #2998: AHCI reports many "Illegal Request" with Intel C602 chipset
- Another status update to observed behavior since behavior has changed again. As of sometime yesterday, the illegal r...
2012-07-30
-
- Assuming we should expect the same functionality as on Solaris files that are removed between snapshots should be lis...
-
- The implementation quirk referenced above, but not described is this:
Given types such as intptr_t, most important... -
- Thanks for reporting this deadlock. I am working on restructuring the way the dp_config_rwlock is used; I'll make su...
-
- Reproducer: "zfs recv & zfs list"
-
- panic[cpu0]/thread=ffffff00d2bc17e0:
Deadlock: cycle in blocking chain
ffffff00024e5a00 genunix:turnstile_blo... -
- Please update this report with:
(b) how much performance could improve if we removed the lock entirely, or changed... -
- Solution: replace rrw_lock with OS rw_lock.
If it brings deadlock, introduce property for rw_lock that "enables" n... -
- ZFS uses it's own rrw_lock.
rrw_lock is not real rw_lock because in internals it is mutex.
Therefore this bring... -
- Currently, one can't delete an uploaded webrev if rsa key is passphrased. It happen because "sftp -b <file>", that is...
2012-07-29
-
- Just cause the build of whatever software to run the GNU version of the command instead? I don't think there's any p...
-
09:09 PM Feature #3039 (Rejected): use new version of xgettext within illumos
- While trying to build GRAMPS ( http://sourceforge.net/projects/gramps/files/ ) the compilation is failing because the...
-
2012-07-28
-
- The swap and free columns are presented in kilobytes by vmstat today. On modern systems with gigabytes of memory this...
2012-07-26
-
- Resolved in r13760 commit:527f645bd884
-
03:27 PM Feature #3035: LZ4 compression support in ZFS and GRUB
- Webrev at http://cr.illumos.org/~webrev/skiselkov/3035/
Source can be pulled from hg staging repo at http://62.65.18... -
- LZ4 is a new high-speed BSD-licensed compression algorithm written by Yann Collet that delivers very high compression...
-
03:24 PM Bug #3036 (Feedback): Update hwdata to current upstream data
- Patches to hwdata DB info for Illumos kernel.
-
- Update hardware data device id data files to recent versions.
Ref: https://hg.openindiana.org/upstream/illumos/illum... -
- The kernel sprintf format string does not currently allow the use of periods.
This increases the difficulty of using... -
- The system is still up and zfs pool is still working ok. No soft/hard/transport errors have been reported. The beha...
-
01:54 AM Bug #1702: D compiler chokes on some bitfields
- This occurs because libdtrace's pseudo-C compiler uses libctf for storing type information, and libctf does not handl...
-
- Resolved in r13758 commit:23432da34147
-
- Resolved in r13758 commit:23432da34147
-
- Resolved in r13758 commit:23432da34147
-
- Resolved in r13758 commit:23432da34147
-
- Resolved in r13758 commit:23432da34147
-
- Resolved in r13758 commit:23432da34147
2012-07-25
-
- Resolved in r13757 commit:2e70a48f5fe2
-
- ...
-
- ...
-
07:19 AM Bug #2930: beadm should sort by creation date
- Working on it...
-
- The DTrace Meta Provider API block comment in <sys/dtrace.h> shows the second argument of dtms_provide_pid and dtms_r...
-
- Resolved in r13755 commit:80fe82480d1e
2012-07-24
-
07:36 PM Feature #1553: ZFS should not trust the layers underneath regarding drive timeouts/failure
- Oh don't get me wrong it should be handled in the driver as well, however ZFS shouldn't trust anything and should alw...
-
12:54 PM Feature #1553: ZFS should not trust the layers underneath regarding drive timeouts/failure
- I think this needs to be dealt with at the driver really, and integrated with FMA, after all other like issues are co...
-
- spam
-
07:13 AM Feature #3029: GM Tech 2
- The Kiwi Bluetooth unit connects to your car via a 6-foot cable with so<a href="http://www.obd4u.fr">obd2</a> bring y...
-
- Bring your own zip-ties or double-sided tape, attach[url=http://www.obd4u.fr]obd2[/url] the Kiwi Bluetooth.Jump to [u...
-
02:58 PM Bug #2995 (Feedback): pcan and pcwl wifi device drivers EOL
- Pending RTI
-
-
-
2012-07-23
-
- Overshoot.
[2] https://bitbucket.org/illumos/illumos-gate/src/7231b684c18b/usr/src/cmd/zfs/zfs_main.c#cl-2483 -
- I think it is @usr/src/cmd/zfs/zfs_main.c@
It uses @nvlist_add_uint32()@ to set id [1], but test for @DATA_TYPE_UI... -
-
- ...
-
- Yuri Pankov wrote:
> there's a column named *USED* in the output, and it should report correct data.
Do you mean ... -
- sham pavman wrote:
> After looking at the concerned code and a few snips from the manpage.. I believe this is the ex... -
05:53 PM Bug #3027: installgrub can segfault when encountering bogus data on disk
- New webrev: http://grumpf.hope-2000.org/illumos-3027-webrev-2/
Changes:
* calculate maximum size of extra data an... -
- Webrev: http://grumpf.hope-2000.org/illumos-3027-webrev/
-
- Installgrub tries to read possibly existing grub stages from the disk before installing the new stages.
For some p... -
10:36 AM Bug #815: Need ipadm(1M) manual page
- Jerry Jelinek (from Joyent) has integrated a likely suitable page into their gate:
https://github.com/joyent/illum... -
- have no dependencies on libz.so on my system with gcc44 build:
root@oi-vm:~# ldd /usr/sbin/lofiadm
libdev... -
12:30 AM Bug #3026 (Resolved): libdtrace should set LD_NOLAZYLOAD=1 to help the pid provider
-
- dt_cg_ptrsize() can leak a register:
185 dt_cg_ptrsize(dt_node_t *dnp, dt_irlist_t *dlp, dt_regset_t *drp,
... -
- casting between types in DTrace can lead to unexpected results; C should be used as the model for how DTrace should b...
-
- Currently you can examine members of dynamic variables:
args[1]->fi_name;
but you cannot copy the whole structure... -
- # dtrace -n 'BEGIN{ @[(int)1, (int)1] = sum(1); @[(short)2, (short)2] = sum(2); exit(0); }'
dtrace: description 'BEG... -
- DTrace should support a mode where output is ordered by time.
Also available in: Atom