Project

General

Profile

Activity

From 2013-10-15 to 2013-11-13

2013-11-13

11:49 PM Bug #4322: ZFS deadlock on dp_config_rwlock
The problem is that dsl_dataset_user_release_impl() calls zfs_unmount_snap() with the dsl_pool_config lock held. As ... Matthew Ahrens
03:26 PM Bug #4322 (Closed): ZFS deadlock on dp_config_rwlock
We encountered a deadlocked zfs and have been able to take a crash dump. From all threads the 3 below seem to be the ... Arne Jansen
10:40 PM Bug #4325: rpc_svc_calls(3nsl): svc_getrpccaller() description should be improved
Okay, svc_getrpccaller() is a macro defined in svc.h. The svc_getrpccaller() function definition in svid_funcs.c is t... Marcel Telka
10:31 PM Bug #4325 (Closed): rpc_svc_calls(3nsl): svc_getrpccaller() description should be improved
In the rpc_svc_calls(3nsl) man page there is this description of svc_getrpccaller():... Marcel Telka
10:32 PM Bug #1872: NULL pointer dereference in iscsit
We have reproduced a variant of this in-house at Nexenta. (BTW, kmem_flags = 0xf didn't help.)
Basically, an iSCS...
Dan McDonald
10:16 PM Bug #4324 (Closed): hosts_access(3): Missing symlinks for hosts_ctl, request_init, and request_set
There are missing symlinks for hosts_ctl.3, request_init.3, and request_set.3 in /usr/share/man/man3/. All those syml... Marcel Telka
10:06 PM Bug #4305: rpc(3nsl): svc_sendreply is in rpc_svc_calls(3nsl)
Similar issue is in rpc_svc_err(3nsl) man page, where it is stated this:... Marcel Telka
09:47 PM Bug #4226 (Resolved): mountd(1m) could run out of file descriptors
... Marcel Telka
05:19 PM Bug #4323 (In Progress): NDMP return the wrong error code in NDMP backup policy
Citing relevant part of the spec:
"NDMP_FILE_NOT_FOUND_ERR During a recover operation, a specified file was no...
Jan Kryl
05:16 PM Bug #4323 (Resolved): NDMP return the wrong error code in NDMP backup policy
NDMP policy has 3 Backup selection paths (with 2 valid and 1 invalid) Backup started with 1st path which was successf... Jan Kryl
02:21 PM Bug #2776: e1000g jumbo frames don't work with 82579LM
Jumbo Frames on 82579LM are working in FreeBSD (10.0-BETA), which is a bit strange, given that we use (nearly) the sa... Yuri Pankov
01:59 PM Bug #4119: Mistake in kwarndtest breaks the build
See #4307. Marcel Telka
01:58 PM Bug #4119 (Closed): Mistake in kwarndtest breaks the build
Marcel Telka
01:53 PM Bug #4307: Variable is used before it is initialized in kwarndtest.c
This is a duplicate of #4119. Next time please search before creating a new issue. Andrew Stormont
10:17 AM Bug #4307 (Resolved): Variable is used before it is initialized in kwarndtest.c
Resolved in:... Yuri Pankov
01:36 PM Bug #4321 (Closed): rpc_svc_calls(3nsl): svc_run() can return
In the rpc_svc_calls(3nsl) man page it is stated that svc_run() never returns:... Marcel Telka
12:09 PM Bug #4320 (Closed): rpc_svc_reg(3nsl): svc_pollfd is modified too
Both xprt_register() and xprt_unregister() modifies both svc_fdset and svc_pollfd, but in the rpc_svc_reg(3nsl) man p... Marcel Telka
05:00 AM Bug #4318 (New): zfs panic: BAD TRAP occurred in module "unix" due to an illegal access to a user address
Latest Illumos git as of 11/11/13, running on an OI 151a8 base.
System with only two pools - an rpool and a single...
Rich Ercolani

2013-11-12

08:31 PM Feature #2583 (Resolved): Add -p (parsable) option to zfs list
... Gordon Ross
07:52 PM Bug #4317: No audio from speakers on some laptops using audiohd
I ran across a fix for this (attached) that was posted by Vladimir Kirillov about two years ago.
http://www.listbox....
Gordon Ross
07:47 PM Bug #4317 (New): No audio from speakers on some laptops using audiohd
On some laptops, the amplifier(s) for the speakers are apparently not getting powered up. The headphone jacks work O... Gordon Ross
04:30 PM Bug #4316: Loud "splat" at boot time on Dell Precision M4500
After a little research, it appears that all the Integrated Devices 92HD{71,75,81} codecs should have the NO_MIXER fl... Gordon Ross
04:25 PM Bug #4316 (New): Loud "splat" at boot time on Dell Precision M4500
The machine uses "audiohd" with codec:
Integrated Devices 92HD81B1C5
Something's not quite right during initializ...
Gordon Ross
12:36 PM Bug #2869: duplicate packets with vnics over aggrs
I am sorry for they delay, but I had some problems I had to solve first. I reproduced the problem and this is the sta... Adam Števko
11:38 AM Bug #4315 (New): mpt_sas panic: BAD TRAP: type=d (#gp General protection)
http://skysrv.pha.jhu.edu/~rercola/vmdump.6 (3.5G)... Rich Ercolani
10:45 AM Bug #4314 (Resolved): mountd(1m): Options are not related to Solaris
The word "Solaris" should be removed here in the mountd(1m) man page because those options are not supported generall... Marcel Telka

2013-11-11

07:16 PM Bug #4312 (New): Dumps to USB storage report success but appear to not actually be written
As the subject says.
On latest Illumos, I was attempting to dump to a non-mpt_sas device to debug an mpt_sas-relat...
Rich Ercolani
06:45 PM Bug #3232 (Feedback): syslogd shouldn't sync after each LOG_KERN line
I've abandoned this bug fix pending further information. Omitting the sync after every write of a system message to ... Gary Mills
06:16 PM Bug #4309: FreeBSD Clang warnings, libzfs
I think, after more public review and comments, that this is "more appropriate" for the code in question. It simply ... sean bruno
06:13 PM Bug #4310: Another mpt_sas panic
Ah, okay. The problem, and underlying meaning of the assertion, is that you can't call taskq_wait from inside a threa... Rich Ercolani
05:35 PM Bug #4310: Another mpt_sas panic
... Rich Ercolani
05:26 PM Bug #4310: Another mpt_sas panic
I have achieved vmdump.
http://skysrv.pha.jhu.edu/~rercola/another_mptsas_vmdump (1.5G)
Debugging to follow.
Rich Ercolani

2013-11-10

11:44 AM Bug #4309: FreeBSD Clang warnings, libzfs
> uint64_t and int64_t should use the printf() specifier
> %ju or %jd for i386/amd64 happiness.
Is there a specif...
Jean-Pierre André
09:30 AM Bug #4310 (Closed): Another mpt_sas panic
Latest git illumos as of today (11/10/13), commit dfdb2065f4.
I was having a problem where my system became unresp...
Rich Ercolani

2013-11-09

01:20 AM Bug #4309 (New): FreeBSD Clang warnings, libzfs
Pretty minor casting/printf things for FreeBSD port of libzfs. More or less, uint64_t and int64_t should use the pri... sean bruno

2013-11-08

04:53 PM Bug #4248 (Resolved): dtrace(1M) should never create DOF with empty probes section
Any userland component that was relying on this behavior is already broken. Any time that this breaks something it in... Robert Mustacchi
03:49 PM Bug #4248 (Feedback): dtrace(1M) should never create DOF with empty probes section
With 54a20ab41aadcb81c53e72fc65886e964e9add59 commit a number of userland components can't be built with Dtrace suppo... Alexander Pyhalov
11:26 AM Bug #4308: svc_userfds reallocation in svc_add_input() is fragile
The similar issue is in _svc_attribute_new_id() in the same source file - usr/src/lib/libnsl/rpc/svc.c. Marcel Telka
10:52 AM Bug #4308 (Resolved): svc_userfds reallocation in svc_add_input() is fragile
Here is how the svc_userfds realloc is implemented in svc_add_input():... Marcel Telka
04:28 AM Bug #4307: Variable is used before it is initialized in kwarndtest.c
Webrev is created for the fix: http://cr.illumos.org/~webrev/usvyatsky/4307/ Ilya Usvyatsky
01:09 AM Bug #4307 (Resolved): Variable is used before it is initialized in kwarndtest.c
gcc -Wuinitialized complains about
https://github.com/illumos/illumos-gate/blob/master/usr/src/cmd/krb5/kwarn/kwarnd...
Ilya Usvyatsky

2013-11-07

08:08 PM Bug #2091: sd fails to power up some disks
A hardware vendor has investigated this and determined that we were sending START STOP UNIT with PC set to 1 and the ... Albert Lee
07:51 PM Bug #4306 (Closed): rpc_svc_calls(3nsl): Messed svc_sendreply/svc_max_pollfd
The rpc_svc_calls(3nsl) man page contains this mess:... Marcel Telka
07:46 PM Bug #4305 (Closed): rpc(3nsl): svc_sendreply is in rpc_svc_calls(3nsl)
The rpc(3nsl) man page says:... Marcel Telka
09:00 AM Feature #4304 (Resolved): fmdump shall emit JSON
It would be most excellent if fmdump(1M) was able to speak the LF-separated streaming JSON format that all the kids a... Robert Mustacchi
08:56 AM Feature #4303 (Resolved): want local zone TCP kstats in the global zone
The "tcp" module provides a kstat of class "mib2" that's particularly interesting. See INTRO-361. There appears to be... Robert Mustacchi
08:51 AM Bug #4302 (Resolved): stack overflow from squeue_drain()
A customer doing load testing induced a panic with the following stack trace:... Robert Mustacchi
08:40 AM Bug #4301 (Resolved): logadm unnecessarily copies historical logs when "-c" is specified
logadm's -c option is defined to "[r]otate the log file by copying it and truncating the original logfile to zero len... Robert Mustacchi
08:40 AM Bug #4300 (Resolved): logadm test suite fails when running rsyslogd
Somewhat amazingly, the logadm test suite has managed to encode a dependency on the implementation of syslogd:
<pr...
Robert Mustacchi
08:40 AM Bug #4299 (Resolved): logadm should not lock log file for entire duration of copy
We experienced an issue on us-beta today where logadm had blocked the marlin agent process for minutes by taking a ma... Robert Mustacchi
08:34 AM Feature #4298 (Resolved): ::printf tab completion
tab complete the ::printf dmod Robert Mustacchi
08:34 AM Feature #4297 (Resolved): ::help tab completion
tab complete the ::help dmod Robert Mustacchi
08:29 AM Bug #4293 (Resolved): mdb typedef load error msg confusing
Resolved in dfdb2065f47c8aecb3ea21b3c7289286092c4a49. Robert Mustacchi
06:28 AM Feature #3968 (Resolved): want FILE_TRUNC event for PORT_SOURCE_FILE
Resolved in 72102e7461c97dc268d21d9dd8f02da45f174acd. Robert Mustacchi
06:28 AM Bug #3930 (Resolved): 'tail -F ...' not resetting the offset of file rotation properly
Resolved in 72102e7461c97dc268d21d9dd8f02da45f174acd. Robert Mustacchi
06:28 AM Bug #3929 (Resolved): `man tail` doesn't mentioned "-F" option
Resolved in 72102e7461c97dc268d21d9dd8f02da45f174acd. Robert Mustacchi
06:28 AM Bug #3928 (Resolved): `tail -f ...` doesn't notice file truncation
Resolved in 72102e7461c97dc268d21d9dd8f02da45f174acd. Robert Mustacchi
01:19 AM Bug #4161: deadlock between zfs_read() and zfs_putpage()
Resolved in b3d32f0ceb59362ba287dcfd6de471e98bfc7fa9. Robert Mustacchi
01:16 AM Bug #4296 (New): dtrace -G mistakingly assigns a function type to symbols
From dtrace-discuss:
Linking vm...
ld: warning: symbol `Universe::_narrow_oop' has differing types:
(fil...
Robert Mustacchi
12:21 AM Bug #4295 (Resolved): libshare sa_get_proto_status sometimes returns unallocated strings
The libshare sa_get_proto_status op expects to have been returned an allocated string from memory that it has to free... Robert Mustacchi

2013-11-06

11:50 PM Bug #4244 (Resolved): spawn fails occasionally with EAGAIN
Resolved in 5a450eee20035a2a426a48161ab3a50e2921878c. Robert Mustacchi
09:50 PM Feature #4294 (Resolved): fopen could support 'x'
C11 adds 'x' as a valid argument to fopen. This matches the traditional usage on other platforms such as the BSDs, OS... Robert Mustacchi
09:36 PM Bug #4293 (Resolved): mdb typedef load error msg confusing
... Robert Mustacchi
09:09 PM Bug #4243: sdev_access() race leads to panic
Resolved in 368fc9412cc48af49761a1b8808d9c4079709391. Robert Mustacchi
07:25 PM Bug #4250 (Resolved): libdtrace should use "F" specifier on fopen()
Robert Mustacchi
07:24 PM Bug #4250: libdtrace should use "F" specifier on fopen()
Resolved in 93ed8d0d4b068b95d0bb50d57bb854df462a8485. Robert Mustacchi
07:24 PM Bug #4251 (Resolved): libdtrace leaks open file handles
Resolved in 93ed8d0d4b068b95d0bb50d57bb854df462a8485. Robert Mustacchi
07:24 PM Bug #4116 (Resolved): kiconv_open(9f): kiconv_close() in the CONTEXT section
Resolved in 393cba3d826b7fd70b8ab083125c4b25c43ee482. Robert Mustacchi
05:37 PM Bug #4245 (Resolved): boot_time kstat for zones should be epoch timestamp of zone boot, not "random" number
Resolved in 86ad481cc8bfbc0155ef221a0fe07ef23db58f68. Robert Mustacchi
05:37 PM Bug #4246 (Resolved): kstat read improperly returned ENOMEM
Resolved in dc9df4786b08572d6032efbd47f727287e691656. Robert Mustacchi
04:10 PM Bug #4237 (Resolved): net_* and hook_* man pages are wrong
Resolved in fe77cc0407fb667ddc04e1a8f2e203bb7b9c80e1. Robert Mustacchi
04:10 PM Feature #4236 (Resolved): Internet Packet Disturber
Resolved in fe77cc0407fb667ddc04e1a8f2e203bb7b9c80e1. Robert Mustacchi
07:12 AM Bug #4290 (Resolved): dlmgmtd is careless with avl_find
Resolved in c6d054cbc999e5c8b9ad1aa01dbb4800b84f06bd. Robert Mustacchi
05:34 AM Bug #4248 (Resolved): dtrace(1M) should never create DOF with empty probes section
Resolved in 54a20ab41aadcb81c53e72fc65886e964e9add59. Robert Mustacchi
05:34 AM Bug #4249 (Resolved): Only probes from the first DTrace object file will be included
Resolved in 54a20ab41aadcb81c53e72fc65886e964e9add59. Robert Mustacchi
03:23 AM Feature #4291 (New): IPD needs "reorder"
A network pathology that IPD does not appear to cover is packet reordering. Often, due to parallel packet processing... Dan McDonald

2013-11-05

11:15 PM Bug #4290 (Resolved): dlmgmtd is careless with avl_find
We found a dump from dlmgmtd. Looking at it Joshua Clulow had the following comments:... Robert Mustacchi
11:01 PM Bug #4289 (Resolved): libnsl: do_accept() shouldn't be recursive
The current implementation of do_accept() in libnsl could call itself recursive when there are more new TCP connectio... Marcel Telka
09:32 PM Bug #4200 (Resolved): prctl(1): Duplicate information in the man page
Resolved in b20613457c9cae13ad78e3ecace17563f4b65134. Robert Mustacchi
08:25 PM Bug #4116: kiconv_open(9f): kiconv_close() in the CONTEXT section
Webrev can be found here:
http://cr.illumos.org/~webrev/bcoddens/illumos-4116/
Bart Coddens
07:45 PM Bug #4116 (In Progress): kiconv_open(9f): kiconv_close() in the CONTEXT section
Easy fix, webrev is coming up Bart Coddens
06:35 PM Bug #4278 (Resolved): read-only lofi devices can never be removed if used
Resolved in a67ac136365c88325c50c83d5b9b7891093cd1da. Robert Mustacchi
06:35 PM Bug #4279 (Resolved): ipmi(7d): Invalid url specified
Resolved in 28bf74a613770b649af5dd9c9b6b99625559d6ac. Robert Mustacchi
06:35 PM Bug #4282 (Resolved): Race between ipmi_close() and kcs_loop() could lead to panic
Resolved in cc944374608a39fb5622959cc9a210fb116539cb. Robert Mustacchi
06:45 AM Bug #1756: iwp missing some Centrino 6205 IDs

>
> Please tell me if there is anything I can do to help someone update the driver to support this device. As a ...
Espen Martinsen

2013-11-04

09:01 PM Bug #4287 (New): FreeBSD Clang reports warning, libctf
Minor warning from clang 3.3:
/home/sbruno/bsd/fbsd_head/cddl/lib/libctf/../../../cddl/contrib/opensolaris/common/ct...
sean bruno
08:30 PM Bug #4286: hermon driver hangs booting
I may be able to provide SSH access to it if really needed. But since it relates to booting I'm not sure how much tha... Luoxia Xue
08:22 PM Bug #4286 (New): hermon driver hangs booting
I have a Mellanox ConnectX-2 VPI MHQH19B-XTR HCA card presented to an OpenIndiana 151a8 VM but it hangs at boot. I t... Luoxia Xue
07:28 PM Bug #4285 (Resolved): preemption leaves a context op behind
Resolved in 7ac89421b5a801d998dd60ee8cd0684bc6e83fcd. Robert Mustacchi
05:48 PM Bug #4285 (Resolved): preemption leaves a context op behind
Ken Harris had long been dealing with an issue regarding FPU corruption on x86 systems on smartos-discuss. He was abl... Robert Mustacchi
04:10 PM Bug #4284 (New): FreeBSD Clang reports warning, libnvpair
Minor warning nit from clang:... sean bruno

2013-11-01

07:10 PM Bug #4233 (Resolved): mptsas topo change buffer overflow
Resolved in 8b1f072cdb7c7ae7c166fc48e2f1988fc5fbb3f4. Robert Mustacchi
04:48 PM Bug #4278 (Pending RTI): read-only lofi devices can never be removed if used
Alexander Eremin
03:54 PM Bug #4264 (Resolved): New features should be added to zpool_get.cfg
Resolved in:... Yuri Pankov
03:54 PM Bug #4276 (Resolved): zfstest should use less common user account than "admin" for ACL tests
Resolved in:... Yuri Pankov
03:53 PM Bug #4263 (Resolved): system/test/zfstest should depend on system/xopen/xcu4
Resolved in:... Yuri Pankov
03:53 PM Bug #4263 (Pending RTI): system/test/zfstest should depend on system/xopen/xcu4
Yuri Pankov
02:36 PM Feature #3525: Persistent L2ARC
arc.c line 5849,
bzero(&dev->l2ad_ublk, sizeof (&dev->l2ad_ublk));
Shouldn't that '&' be removed?
Yuxuan Shui

2013-10-31

08:42 PM Bug #4282 (Resolved): Race between ipmi_close() and kcs_loop() could lead to panic
ipmi_complete_request() could access freed memory after the race:... Marcel Telka
07:09 PM Bug #4280 (Rejected): sys/ipmi.h should be in system/header package
Marcel Telka
07:09 PM Bug #4280 (Feedback): sys/ipmi.h should be in system/header package
Okay, that makes sense. Marcel Telka
06:07 PM Bug #4280: sys/ipmi.h should be in system/header package
In general, we've been trying to move stuff *out* of system/header, into the packages that actually deliver the piece... Rich Lowe
04:36 PM Bug #4280 (In Progress): sys/ipmi.h should be in system/header package
Marcel Telka
04:36 PM Bug #4280 (Feedback): sys/ipmi.h should be in system/header package
I agree. Especially when I found that /usr/include/libipmi.h is in system/header too. Marcel Telka
02:24 PM Bug #4280: sys/ipmi.h should be in system/header package
Common doesn't mean correct, I actually liked the idea. Yuri Pankov
02:22 PM Bug #4280 (Rejected): sys/ipmi.h should be in system/header package
Marcel Telka
02:21 PM Bug #4280 (Feedback): sys/ipmi.h should be in system/header package
Further analysis showed that this is common. For example /usr/include/sys/stmf.h is in storage/stmf package. So, I'll... Marcel Telka
01:56 PM Bug #4280 (Rejected): sys/ipmi.h should be in system/header package
The sys/ipmi.h file should be moved from driver/ipmi package to system/header package. The rationale is simple: I wan... Marcel Telka
07:08 PM Bug #4200: prctl(1): Duplicate information in the man page
webrev can be found here:
http://cr.illumos.org/~webrev/bcoddens/illumos-4200/
Bart Coddens
06:34 PM Bug #4200 (In Progress): prctl(1): Duplicate information in the man page
I will take this one Bart Coddens
04:10 PM Bug #4281: nhm and nhmex are dangerous
This is the same issue I hit on Athlon64 -- there, intel_nhmex entirely buggers the APIC.
A fix for that bug is (i...
Rich Lowe
03:09 PM Bug #4281 (New): nhm and nhmex are dangerous
Both the intel_nhm and intel_nhmex drivers crash on some platforms. In particular, I have an Ivy Bridge based platfo... Garrett D'Amore
01:08 PM Bug #4279 (Resolved): ipmi(7d): Invalid url specified
There is double "http" here in the ipmi(7d) man page:... Marcel Telka
09:18 AM Bug #4278 (In Progress): read-only lofi devices can never be removed if used
Alexander Eremin

2013-10-30

08:16 PM Bug #4239 (Resolved): rpc_svc_calls(3nsl): Typo "service service"
... Marcel Telka
07:58 PM Bug #4239 (Pending RTI): rpc_svc_calls(3nsl): Typo "service service"
RTI is in progress Bart Coddens
07:44 PM Bug #4278 (Resolved): read-only lofi devices can never be removed if used
... Rich Lowe

2013-10-28

07:12 PM Bug #4193 (Resolved): ls --color with filename arguments does not colorize
commit 71ef9ec8a14e5d11c2ac4a72fb5717a330206444
Author: Lauri Tirkkonen <lotheac@iki.fi>
Date: Wed Oct 23 02:49:1...
Dan McDonald
07:07 PM Bug #4193: ls --color with filename arguments does not colorize
Dan McDonald wrote:
> Can you please also update the bug report with a quick sentence on why moving
> the initializ...
Lauri Tirkkonen
06:54 PM Bug #4263: system/test/zfstest should depend on system/xopen/xcu4
Note that I'm not really sure about adding explicit dependency here, I just didn't have system/xopen/xcu4 installed a... Yuri Pankov
06:46 PM Bug #4263 (In Progress): system/test/zfstest should depend on system/xopen/xcu4
Easy fix, I will take this one Bart Coddens
06:33 PM Bug #4239: rpc_svc_calls(3nsl): Typo "service service"
Webrev can be found here:
http://cr.illumos.org/~webrev/bcoddens/illumos-4239/
Bart Coddens
05:56 PM Bug #4239 (In Progress): rpc_svc_calls(3nsl): Typo "service service"
Easy fix, I will take this one Bart Coddens
01:35 PM Bug #4276 (Resolved): zfstest should use less common user account than "admin" for ACL tests
"admin" user account is not that uncommon, and its existence is causing zftest's ACL tests to fail. We could use some... Yuri Pankov

2013-10-26

03:16 PM Bug #1762 (In Progress): Syslogd man page: missing reference.
I'll fix this one as part of two other syslog fixes.
Gary Mills
03:15 PM Bug #3232 (In Progress): syslogd shouldn't sync after each LOG_KERN line
My strategy is to detect bursts of kernel messages by calling poll() repeatedly with zero timeout until there are no ... Gary Mills
03:10 PM Bug #4211 (In Progress): Some syslog facility names and symbols are missing
Where there were no conflicts, I decided to use the names and symbols already defined in BSD Unix. In the case of fa... Gary Mills

2013-10-25

01:50 PM Bug #4088 (Resolved): use after free in arc_release()
... Marcel Telka
01:49 PM Bug #4089 (Resolved): NULL pointer dereference in arc_read()
... Marcel Telka
01:33 PM Bug #4212 (Resolved): /etc/logindevperm contains two identical lines
Marcel Telka
01:33 PM Bug #4212 (Pending RTI): /etc/logindevperm contains two identical lines
Marcel Telka
01:32 PM Bug #4212: /etc/logindevperm contains two identical lines
... Marcel Telka

2013-10-24

09:52 PM Bug #4272 (New): NFSv4 server: OPEN/CLAIM_DELEGATE_CUR ignores directory and file
The OPEN/CLAIM_DELEGATE_CUR implementation in rfs4_do_opendelcur() ignores directory (in CURRENT_FH) and file name pa... Marcel Telka
08:56 PM Bug #4271 (New): NFSv4 server: delegations should be recalled before the nbl_conflict() call
Delegation support in NFSv4 allows the NFS server to delegate some responsibilities to NFS clients. If such delegatio... Marcel Telka
07:37 PM Bug #4270 (Resolved): ld(1) argument error reporting is still pretty bad
There are still several problems with the error reporting of ld(1).
The two I most recently ran into are:...
Rich Lowe
07:28 PM Bug #4227 (In Progress): ld --library-path is translated to -l-path, not -L
Rich Lowe
02:57 PM Bug #4267 (Closed): RUNPATH missing from c++ executables
This is because of Keith's changes for #3915 Add adjuncts support to the build, which was accompanied by a flag day m... Rich Lowe
02:00 PM Bug #4267 (Closed): RUNPATH missing from c++ executables
When built with gcc, the RUNPATH is missing from executables compiled from c++ source. The errors in mail_msg are li... Gary Mills
02:55 PM Bug #4268 (New): zfs, genuinx or unix NULL pointer dereference
My Openindiana oi_148 server has started to reboot itself. Can anyone spread light if this is a issue already fixed, ... John McEntee
02:24 PM Bug #3790 (In Progress): Some Makefiles have LINT.c without LDLIBS
This bug has been fixed at least for the case illustrated above by additions to usr/src/cmd/sgs/error/i386/Makefile a... Gary Mills
12:57 PM Bug #4266 (Resolved): remove Makefile*.shared nonsense in uts
AFIAK, the .shared makefile constructs appeared to support building of closed source code. Since we do not have acce... Josef Sipek
12:53 PM Bug #4265 (Resolved): remove INTERNAL_RELEASE_BUILD
The INTERNAL_RELEASE_BUILD flag caused a release build to have a debug build-like version string in the linker compon... Josef Sipek
02:28 AM Bug #4234: PATH=/usr/gnu/bin:$PATH breaks things
There was also a suggestion that we simply remove these files. That's an interesting solution that avoids potential a... Gordon Ross
02:07 AM Bug #4029 (Resolved): remove tonic build bits
Resolved in c0e7977 Rich Lowe
02:07 AM Bug #4028 (Resolved): remove CLOSED_IS_PRESENT
Resolved in c0e7977 Rich Lowe
02:07 AM Bug #4027 (Resolved): remove CLOSED_BUILD
Resolved in c0e7977 Rich Lowe
02:07 AM Bug #4030 (Resolved): remove useless nightly/bldenv options
Resolved in 5110237 Rich Lowe
02:07 AM Bug #4034 (Resolved): 'dmake all' fails for DTrace demos
Resolved in 90c1434 Rich Lowe
02:07 AM Bug #4215 (Resolved): pkcs11_tpm botches shared library initialization, murders its friends
Resolved in 9048537 Rich Lowe

2013-10-23

09:20 PM Bug #4264 (Resolved): New features should be added to zpool_get.cfg
Currently, zfs-tests/tests/functional/cli_root/zpool_get/zpool_get_002_pos fails due to missing new features in its .... Yuri Pankov
07:35 PM Bug #4263 (Resolved): system/test/zfstest should depend on system/xopen/xcu4
So that /usr/xpg4/bin/sh is available. Or, probably, chg_usr_exec should do better at error reporting, I had a hard t... Yuri Pankov
01:05 PM Bug #4262 (New): NFSv4 server: DELEGPURGE violates RFC
The NFSv4 operation DELEGPURGE (as implemented by our server) violates RFC 7530 because it always returns NFS4ERR_INV... Marcel Telka
08:09 AM Bug #1756 (Feedback): iwp missing some Centrino 6205 IDs
Hi,
I'm installing OI 151.a8 from scratch on a new Lenovo PC with this
builtin wireless:
pci bus 0x0003 ca...
Espen Martinsen

2013-10-22

10:13 PM Bug #4226 (Pending RTI): mountd(1m) could run out of file descriptors
Marcel Telka
11:36 AM Bug #4259 (New): NFSv4 server should handle the server edge conditions
The NFSv4 server should properly handle the server edge conditions as described in section _9.6.3.4.3. Handling Serve... Marcel Telka
02:50 AM Bug #4257 (Closed): no manpage for ipadm
Duplicate of #815 Rich Lowe
01:06 AM Bug #4257 (Closed): no manpage for ipadm
@> man ipadm
No manual entry for ipadm.
[18:04:23] ops@sidekiq-saves002.prod [~]
> ipadm --help
ipadm: unknown su...
Blake Irvin
01:00 AM Feature #4256 (New): want '-s' flag for svcadm restart so that we know whether or not restart method succeeded
should behave more or less like 'svcadm start/stop -s <fmri>' Blake Irvin
12:59 AM Feature #4255 (New): want flag for svcadm to allow operator to override method timeout
this would be extremely useful for emergencies when a service needs to be shut down immediately Blake Irvin

2013-10-21

12:39 PM Bug #4252 (Rejected): kernel panic with zfs send/receive while upgrading to a vdev with larger ashift
Marcel Telka
08:31 AM Bug #4252 (Feedback): kernel panic with zfs send/receive while upgrading to a vdev with larger ashift
sorry, duplicate of #4188 Richard PALO
06:41 AM Bug #4252 (Rejected): kernel panic with zfs send/receive while upgrading to a vdev with larger ashift
Too much downtime... after upgrading sd.conf with http://wiki.illumos.org/display/illumos/List+of+sd-config-list+ent... Richard PALO
06:22 AM Bug #4230: zpool import panics system
should be fixed by the following change.
I read the latest source code, but the code actually run is a few days older...
Zhiwen Zheng
05:53 AM Bug #4230: zpool import panics system
spa_import::dis... Zhiwen Zheng
05:34 AM Bug #4230: zpool import panics system
it is doing verbatim import.
::print -aT proc_t...
Zhiwen Zheng
03:37 AM Bug #4230: zpool import panics system
nvlist_t *spa_config = 0xffffff033ed88910... Zhiwen Zheng
02:48 AM Bug #4230: zpool import panics system
... Zhiwen Zheng
02:38 AM Bug #4230: zpool import panics system
... Zhiwen Zheng
02:36 AM Bug #4230: zpool import panics system
... Zhiwen Zheng

2013-10-20

06:24 PM Bug #4251 (Resolved): libdtrace leaks open file handles
From dt_load_libs_dir:... Robert Mustacchi
06:24 PM Bug #4250 (Resolved): libdtrace should use "F" specifier on fopen()
While this was fixed in dtrace_freopen(), the fopen() calls in dt_load_libs_dir() and dt_load_libs_sort() don't use t... Robert Mustacchi
06:17 PM Bug #4249 (Resolved): Only probes from the first DTrace object file will be included
From the original Joyent bug report:
When trying to link together multiple object files generated by DTrace only t...
Robert Mustacchi
06:17 PM Bug #4248 (Resolved): dtrace(1M) should never create DOF with empty probes section
Previously you could use dtrace -G and entirely mess up the invocation and end up creating zero probe sites. This lea... Robert Mustacchi
06:08 PM Bug #4247 (New): Reduce thread migrations caused by pg_cmt_can_migrate()
Based on extensive analysis by Brendan Gregg we found that we were often doing many more steals to other CPUs that wo... Robert Mustacchi
12:08 PM Bug #4229: mdb hangs on exit when long umem cache names exist
Here are the possible solutions...
# update every user of mdb_nv_insert to check the length and strdup as necessar...
Josef Sipek
12:32 AM Bug #4246 (Resolved): kstat read improperly returned ENOMEM
Summarized from the Joyent bug report:... Robert Mustacchi
12:27 AM Bug #4245 (Resolved): boot_time kstat for zones should be epoch timestamp of zone boot, not "random" number
From the Joyent bug report:
Currently the zone_boot_time is calculated by:...
Robert Mustacchi
12:21 AM Bug #4244 (Resolved): spawn fails occasionally with EAGAIN
We have seen cases – to date, quite rare – whereby a spawn in the global zone fails with EAGAIN, throwing the followi... Robert Mustacchi
12:13 AM Bug #4243 (Resolved): sdev_access() race leads to panic
From the Joyent bug report:
We died in sdev_unlocked_access() as follows:...
Robert Mustacchi

2013-10-19

10:10 PM Bug #4242 (Closed): file rename event fires before the rename happens
From Joyent's OS-2557:
So we're basically just doing a check here that after we got a 'rename' event for our file,...
Robert Mustacchi
06:36 PM Bug #4241 (In Progress): e1000g races in PHY accesses
The e1000g driver has a few races in the way it accesses it PHY registers. In particular, multiple kstat readers, an... Garrett D'Amore
05:23 PM Bug #4229: mdb hangs on exit when long umem cache names exist
Ok, so summary so far...
libumem has cache names. The names can be up to 32 bytes long (including the trailing nu...
Josef Sipek
07:40 AM Bug #4240 (Closed): lofs(7fs): Missing quotes
There are missing quotes around lofs here in lofs(7fs) man page:... Marcel Telka
12:47 AM Bug #4217 (Resolved): Writes to MSI-X table should be DWords, not QWords
commit 8181b438236881e6d31f7155101c85c011a9b6bb
Author: Garrett D'Amore <garrett@openindiana.(none)>
Date: Wed Oc...
Garrett D'Amore
12:24 AM Bug #4239 (Resolved): rpc_svc_calls(3nsl): Typo "service service"
There is double "service" here in the rpc_svc_calls(3nsl) man page:... Marcel Telka

2013-10-18

10:01 PM Feature #4238 (New): hyprlofs
hyprlofs is a filesystem that is a hybrid of lofs and tmpfs. Conceptually it
is similar to lofs, but unlike lofs, wh...
Robert Mustacchi
09:47 PM Bug #4237 (Resolved): net_* and hook_* man pages are wrong
This ticket covers issues with the following manual pages:
* net_hook_register
* net_protocol_lookup
* net_protc...
Robert Mustacchi
09:47 PM Feature #4236 (Resolved): Internet Packet Disturber
We would like to be able to cause a given netstack to engage in pathological behavior that mimics what happens when g... Robert Mustacchi
06:28 PM Bug #4229: mdb hangs on exit when long umem cache names exist
Quick update:
Changing the MDB_NV_NAMELEN #define from 31 to 32 makes mdb happy. In other words, the issue is wit...
Josef Sipek
12:51 AM Bug #4229 (Resolved): mdb hangs on exit when long umem cache names exist
libumem allows caches to have an ASCII name - up to 31 chars. If a 31-char long name is used, mdb will hang on exit.... Josef Sipek
04:43 PM Bug #4212: /etc/logindevperm contains two identical lines
Webrev can be found here:
http://cr.illumos.org/~webrev/bcoddens/illumos-4212/
Bart Coddens
04:27 PM Bug #4235 (Resolved): beadm(1M) mount options undocumented in man page
beadm man page doesn't contain lines about mount options:
beadm mount [-s ro|rw] [-v] beName [mountpoint]
Alexander Eremin
03:25 PM Bug #4234: PATH=/usr/gnu/bin:$PATH breaks things
We've lived with this annoyance long enough. The fix is easy. See
usr/src/cmd/nsadmin/dot-profile.sh
and take o...
Gordon Ross
03:23 PM Bug #4234 (New): PATH=/usr/gnu/bin:$PATH breaks things
See also:
https://defect.opensolaris.org/bz/show_bug.cgi?id=576
David Comay <david.comay@oracle.com> changed:
...
Gordon Ross
03:12 PM Bug #4233: mptsas topo change buffer overflow
Webrev: http://cr.illumos.org/~webrev/hans/illumos-4233-webrev/
To fix this, I split mptsas_smhba_set_phy_props()....
Hans Rosenfeld
03:02 PM Bug #4233 (Resolved): mptsas topo change buffer overflow
There is a nasty little buffer overflow in the SMHBA code in mpt_sas.
When a directly attached device is plugged i...
Hans Rosenfeld
01:28 PM Bug #4198 (Resolved): Assertion failed in nlm_slock_wait()
4198 Assertion failed in nlm_slock_wait()
Reviewed by: Garrett D'Amore <garrett@damore.org>
Reviewed by...
Garrett D'Amore
06:06 AM Feature #1023: nv_sata support for NVIDIA MCP61
I glossed over the first line of my extract, indicative I wasn't getting dma.
After applying Jon's update of Albert'...
Richard PALO
03:32 AM Bug #4230: zpool import panics system
... Zhiwen Zheng
03:22 AM Bug #4230 (Closed): zpool import panics system
ffffff00107e7790 unix:die+e9 ()
ffffff00107e78a0 unix:trap+13c0 ()
ffffff00107e78b0 unix:cmntrap+1ca () ...
Zhiwen Zheng
12:00 AM Bug #3669 (Resolved): for local auditors, LA_FLG_BINDTO does not audit all symbol bindings
Resolved in 38f4bddda7216cf3550c325e8cabe56d08a2bce9. Robert Mustacchi

2013-10-17

10:50 PM Bug #4228 (Closed): chmod(1): SFI NFSv4 ACL inheritance flags are not documented
The successful_access (S), failed_access (F), and inherited (I) NFSv4 ACL inheritance flags are supported by chmod(1)... Marcel Telka
08:44 PM Bug #4227 (Resolved): ld --library-path is translated to -l-path, not -L
... Rich Lowe
07:34 PM Bug #4226 (Resolved): mountd(1m) could run out of file descriptors
*The main issue*
mountd(1m) is running with limit of 256 file descriptors:...
Marcel Telka
05:56 PM Bug #4225 (In Progress): ehci can hang interminably trying to read PCI capabilities
Garrett D'Amore
05:56 PM Bug #4225 (In Progress): ehci can hang interminably trying to read PCI capabilities
The ehci driver has an unbounded loop trying to read extended PCI capabilities. For properly initialized hardware, t... Garrett D'Amore
04:35 PM Bug #4212 (In Progress): /etc/logindevperm contains two identical lines
Easy Fix, I will take this one Bart Coddens
02:39 PM Bug #4224 (New): Possible null pointer dereference in mptsas_hash_add()
In a case the kmem_zalloc() returns NULL (please note KM_NOSLEEP) we will panic at line 15388 in mptsas_hash_add():
...
Marcel Telka
02:22 PM Feature #2583 (In Progress): Add -p (parsable) option to zfs list
#3993 needs more work. Yuri Pankov
02:21 PM Feature #4223 (New): tolerate world-readable closed_bins
I like to use a shared (read-only) copy of the closed_bins on our build machine, shared by many user build workspaces... Gordon Ross
02:12 PM Feature #4222: Factor out maxjobs default stuff from env file(s)
What about bldenv? Yuri Pankov
02:08 PM Feature #4222 (New): Factor out maxjobs default stuff from env file(s)
There's now a (somewhat large) function in the sample env file
($SRC/tools/env/illumos.sh) that computes a default v...
Gordon Ross
02:06 PM Feature #4221 (New): Helper script for building under Jenkins
Chris Siden and Delphix (and others?) came up with a nice little script to check the mail_msg file after a build to m... Gordon Ross
05:41 AM Bug #4220 (Resolved): beadm mount of old BE with zones fails "Read only filesystem"
I have an old BE, it has associated zones which I have since deleted.
If I try to mount it:...
Rich Lowe
05:26 AM Bug #4198: Assertion failed in nlm_slock_wait()
From condvar(9f):... Marcel Telka
02:19 AM Bug #4219 (New): zfs: allocating allocated segment

panic[cpu1]/thread=ffffff04eae2d120:
zfs: allocating allocated segment(offset=182868315136 size=36864)
fffff...
Zhiwen Zheng

2013-10-16

11:08 PM Feature #4218 (New): Tunable ARC buf hash sizing
At present the ARC buffer hash table sizing is hard-coded to use 128k for every 1GB of physical memory installed. Thi... Sašo Kiselkov
09:55 PM Bug #4217: Writes to MSI-X table should be DWords, not QWords
Garrett D'Amore wrote:
> That description got bungled by the tracking system:
The Description is editable. I just...
Marcel Telka
09:27 PM Bug #4217: Writes to MSI-X table should be DWords, not QWords
That description got bungled by the tracking system:
The folks that developed the Intel Communications Chipset 89xx ...
Garrett D'Amore
09:27 PM Bug #4217 (Resolved): Writes to MSI-X table should be DWords, not QWords
Accesses to MSI-X Table entries should be 32-bits (DWords) at a time. However, we use QWords to access the high and ... Garrett D'Amore
08:46 PM Bug #4198 (Pending RTI): Assertion failed in nlm_slock_wait()
Marcel Telka
08:12 PM Bug #4216 (New): libraries should not manipulate the system log
At present, pkcs11_tpm makes use of the system log facilities. This is a terrible idea from a shared library, especi... Rich Lowe
08:03 PM Bug #4215 (Resolved): pkcs11_tpm botches shared library initialization, murders its friends
pkcs11_tpm calls _init() in C_GetFunctionList(). This causes us to call all of our initializers twice, once on libra... Rich Lowe
06:06 PM Bug #4214 (New): zfs destroy -Rnv doesn't show recursion
Consider I have two filesystems:
rpool/export/home/danmcd/ws/stash-nza@stash-crunch 198K - 848M -
r...
Dan McDonald
01:20 PM Bug #4141 (In Progress): deleg_reopen() can leak osp->os_ref_count
Currently, we have a fix under testing. We believe it's not so easy to reproduce this bug. S. Klinkert
06:53 AM Feature #3890: New deduplication hash algorithm
Hello that is not true,
actually fletcher4 is only usable for checksumming not for deduplication.
Today the onl...
Anthony Glidic

2013-10-15

10:42 PM Bug #4213 (Closed): Missing spaces in man pages
There is missing space before "will" here in mutex_init(3c):... Marcel Telka
08:09 PM Bug #4212 (Resolved): /etc/logindevperm contains two identical lines
The /etc/logindevperm file contains two identical lines for /dev/mouse and /dev/kbd:... Marcel Telka
04:40 PM Bug #4188: assertion failed in dmu_tx_hold_free(): dn_datablkshift != 0
Igor Kozhukhov wrote:
> Matthew Ahrens wrote:
> > Great to hear that that patch is working. I am working to get it...
Matthew Ahrens
04:25 PM Bug #4188: assertion failed in dmu_tx_hold_free(): dn_datablkshift != 0
Matthew Ahrens wrote:
> Great to hear that that patch is working. I am working to get it in illumos, hopefully will...
Igor Kozhukhov
03:59 PM Bug #4188: assertion failed in dmu_tx_hold_free(): dn_datablkshift != 0
Great to hear that that patch is working. I am working to get it in illumos, hopefully will be in later this week or... Matthew Ahrens
02:16 PM Bug #4188: assertion failed in dmu_tx_hold_free(): dn_datablkshift != 0
We are running with a trivial patch of just removing the assertion and are not seeing any problems could be considere... Andriy Gapon
03:45 PM Bug #4211 (Resolved): Some syslog facility names and symbols are missing
The syslog facility numbers 9, 10, 11, 12, and 14 are missing from /usr/include/sys/syslog.h . They have no symbols ... Gary Mills
12:52 PM Bug #4123: Locks should not be held across the call to ddi_periodic_delete(9f)
Happened again. It seems it is worth fixing... Marcel Telka
11:41 AM Bug #1681: support for BUILD64=# no longer works
An example of the problem:... Andrew Stormont
07:40 AM Bug #1666: libwrap is 32bit only
Ok. Got it. Alexander Pyhalov
07:31 AM Bug #1666: libwrap is 32bit only
I'm not sure, but man strerror says:
Upon successful completion, strerror() returns a pointer to
the ge...
Alexander Pyhalov
12:31 AM Feature #3964 (Resolved): L2ARC should always compress metadata buffers
commit e4be62a2b74a8f09bb669217a1a39eee069b13a1
Author: Saso Kiselkov <skiselkov.ml@gmail.com>
Date: Mon Oct 14 1...
Dan McDonald
 

Also available in: Atom