Findings
010db139 -- SEGV #1 icns_read_be (size=2, inp=0x661000, outp=<synthetic pointer>) at icns_io.c:51
3454826d -- SEGV #1 icns_read_le (size=2, inp=<optimized out>, outp=<synthetic pointer>) at icns_io.c:101
Reported to maintainer Mathew Eis March 14, 2015, 9:43 p.m.
Saturday, March 14, 2015
Sunday, March 8, 2015
Fuzzing Update
freeXL:
Re: 9f74b0e8, c9be2aa7, d7273f72, 6889d18b
Closed with patch.
https://www.gaia-gis.it/fossil/freexl/fdiff?v1=2e167b337481dda3&v2=61618ce51a9b0c15&sbs=1
html2text:
Re: 8afd955e [Crash, SEGV ] malloc in Area.C:223
Response received from vendor, wontfix.
exifprobe:
Re: 3a41cb1c, 6bb4db60
Closed with patch. 3a41cb1c resolved in this patch. 6bb4db60 resolved in this patch.
Re: ab50ccf7
Closed in latest git checkout. Asking vendor for patch details.
fig2xsd:
Re: 485f75d6, bd644d4d
Closed with vendor release of 0.22
http://prdownloads.sourceforge.net/fig2sxd/fig2sxd_0.22.orig.tar.gz?download
gif2png:
Re: 0cd10b14
Vendor working on fix.
hp2xx:
Submitted new 0ad69d36, 24dfd181, 27eac1b7, 2b9f1f19, 3503093d, 3a5b6ff4, 3b2a6ca6, 40dd99c6, 42e898d9, 4fa145b4, 5570ae2a, 57dc6ada, 7c503fc7, 7d580ece, 87aebb39, 8d2e9b05, 9071d6d8, 91fe0b08, 92cee751, 99f38038, a921cb74, b1ee740e, b3445f4e, b6bd0cc0, b9eff3a0, bbb8391b, c3e9eae3, c7f0611a, cb3d04f9, d34f9134, d6ce68b4, dccd6129, ea388b76, f2315602, f7cede9e on 3/8/2015
giftrans:
Submitted new a4ac94f0 to Debian QA team on 3/8/2015, Debian bug ID 780079;
Re: 9f74b0e8, c9be2aa7, d7273f72, 6889d18b
Closed with patch.
https://www.gaia-gis.it/fossil/freexl/fdiff?v1=2e167b337481dda3&v2=61618ce51a9b0c15&sbs=1
html2text:
Re: 8afd955e [Crash, SEGV ] malloc in Area.C:223
Response received from vendor, wontfix.
exifprobe:
Re: 3a41cb1c, 6bb4db60
Closed with patch. 3a41cb1c resolved in this patch. 6bb4db60 resolved in this patch.
Re: ab50ccf7
Closed in latest git checkout. Asking vendor for patch details.
fig2xsd:
Re: 485f75d6, bd644d4d
Closed with vendor release of 0.22
http://prdownloads.sourceforge.net/fig2sxd/fig2sxd_0.22.orig.tar.gz?download
gif2png:
Re: 0cd10b14
Vendor working on fix.
hp2xx:
Submitted new 0ad69d36, 24dfd181, 27eac1b7, 2b9f1f19, 3503093d, 3a5b6ff4, 3b2a6ca6, 40dd99c6, 42e898d9, 4fa145b4, 5570ae2a, 57dc6ada, 7c503fc7, 7d580ece, 87aebb39, 8d2e9b05, 9071d6d8, 91fe0b08, 92cee751, 99f38038, a921cb74, b1ee740e, b3445f4e, b6bd0cc0, b9eff3a0, bbb8391b, c3e9eae3, c7f0611a, cb3d04f9, d34f9134, d6ce68b4, dccd6129, ea388b76, f2315602, f7cede9e on 3/8/2015
giftrans:
Submitted new a4ac94f0 to Debian QA team on 3/8/2015, Debian bug ID 780079;
Sunday, January 25, 2015
Fuzzing Update
ImageMagick:
Mutliple CPU DOS
http://www.imagemagick.org/discourse-server/viewtopic.php?f=3&t=26929
http://www.imagemagick.org/discourse-server/viewtopic.php?f=3&t=26931
http://www.imagemagick.org/discourse-server/viewtopic.php?f=3&t=26932
http://www.imagemagick.org/discourse-server/viewtopic.php?f=3&t=26933
GraphicsMagick:
Cineon stack smashing and Fits assertion errors 1/24
http://www.graphicsmagick.org/Changelog.html
Mozjpeg:
SIGFPE on jpeg
https://github.com/mozilla/mozjpeg/issues/153
Mutliple CPU DOS
http://www.imagemagick.org/discourse-server/viewtopic.php?f=3&t=26929
http://www.imagemagick.org/discourse-server/viewtopic.php?f=3&t=26931
http://www.imagemagick.org/discourse-server/viewtopic.php?f=3&t=26932
http://www.imagemagick.org/discourse-server/viewtopic.php?f=3&t=26933
GraphicsMagick:
Cineon stack smashing and Fits assertion errors 1/24
http://www.graphicsmagick.org/Changelog.html
Mozjpeg:
SIGFPE on jpeg
https://github.com/mozilla/mozjpeg/issues/153
Saturday, December 29, 2012
sysidcfg in solaris 10 shared-ip zones
If you've tried to put a sysidcfg in shared-ip zones with multiple network interfaces, you've probably experienced some headaches.
I found that if I specified each interface like you would in a normal sysidcfg, it would drop to interactive input.
When I was just specifying one interface with network_interface=PRIMARY{etc etc}, everything worked fine. When I would switch to network_interface=e1000g0{primary hostname=server.local} , it would drop to interactive again.
The secret was just to use PRIMARY no matter how many interfaces I was configuring.
After some tinkering I was able to drop the whole network_interface{} stanza down to:
network_interface=PRIMARY{hostname=hostname.local}
I found that if I specified each interface like you would in a normal sysidcfg, it would drop to interactive input.
When I was just specifying one interface with network_interface=PRIMARY{etc etc}, everything worked fine. When I would switch to network_interface=e1000g0{primary hostname=server.local} , it would drop to interactive again.
The secret was just to use PRIMARY no matter how many interfaces I was configuring.
After some tinkering I was able to drop the whole network_interface{} stanza down to:
network_interface=PRIMARY{hostname=hostname.local}
Friday, November 30, 2012
Verify backups before implementing changes
Going to make a change to a machine?
Verify that you have good colds before you go through.
Almost every change or upgrade plan I have starts off with:
If that backup isn't good, either take a good backup immediately before preceding or reschedule your implementation for a window after the next full backup.
I've had this save me a time or two. A previously bulletproof backup decided to cook off the week of one of my upgrades.
Verify that you have good colds before you go through.
Almost every change or upgrade plan I have starts off with:
- Day after colds: Verify backups for hosts {a,b,c,d,e,f,g,h}
- Day of implementation: Verify backups for hosts {a,b,c,d,e,f,g,h}
If that backup isn't good, either take a good backup immediately before preceding or reschedule your implementation for a window after the next full backup.
I've had this save me a time or two. A previously bulletproof backup decided to cook off the week of one of my upgrades.
Friday, March 11, 2011
Oracle 10g on ZFS IOPS
I was reviewing a machine of ours that people had been complaining about. The disks were horribly busy and it made it difficult to tell what was hurting the performance in Oracle Apps. zpool iostat for the DB pool was constantly running about 4,000-6,000 read OPs at up to 65MB/sec, which seemed off. The ARC prefetch misses were really high and it occurred to me that ARC was prefetching at all. After disabling file-level prefetch the zpool iostat numbers dropped to 0-150 read OPs immediately. I really thought I had broken it, but I logged in to the apps and tested things out, everything was snappy.
Many thanks to ZFS Evil Tuning guide.
Many thanks to ZFS Evil Tuning guide.
Subscribe to:
Posts (Atom)