scientificlinuxforum.org QR code
Scientific Linux Forum.org



  Reply to this topicStart new topicStart Poll

> SL 6.1 x86_64 Skype Crashes, SL 6.1 x86_64 Skype Crashes
mgmartin
 Posted: Sep 13 2011, 08:41 PM
Quote Post


SLF Newbie


Group: Members
Posts: 2
Member No.: 834
Joined: 13-September 11









I'm wondering if anyone is running Skype well on 6.1/64 bit. I recently replaced Fedora with SL on my desktop, and everything runs great, except for Skype.

Surprisingly, Skype ran very well for me on Fedora. On a new SL install, I have a hard time getting it to start. I generally see this error:

CODE
*** glibc detected *** ./skype: malloc(): memory corruption: 0x0a6cb250 ***


An strace reveals this:
CODE

fstat64(11, {st_mode=S_IFREG|0644, st_size=509726, ...}) = 0
mmap2(NULL, 512000, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0xfffffffff70bc000
fstat64(11, {st_mode=S_IFREG|0644, st_size=509726, ...}) = 0
mmap2(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0xfffffffff7775000
read(11, "# UTF-8 (Unicode) compose sequen"..., 4096) = 4096
open("/usr/share/X11/XKeysymDB", O_RDONLY) = 12
fstat64(12, {st_mode=S_IFREG|0644, st_size=9011, ...}) = 0
read(12, "! $Xorg: XKeysymDB,v 1.3 2000/08"..., 4294967295) = 9011
close(12)                               = 0
open("/dev/tty", O_RDWR|O_NOCTTY|O_NONBLOCK) = 12
writev(12, [{"*** glibc detected *** ", 23}, {"./skype", 7}, {": ", 2}, {"malloc(): memory corruption", 27}, {": 0x", 4}, {"0aa34e58", 8}, {" ***\n", 5}], 7*** glibc detected *** ./skype: malloc(): memory corruption: 0x0aa34e58 ***
) = 76


Further investigation shows if I remove /usr/share/X11/XKeysymDB skype starts up and works fairly well, until I send/receive video. The video works for a bit, then blows up. The mere existance of an empty XKeysymDB will cause the error on start up, and Skype will not start up.

Here is an strace after launching with no XKeysymDB and then starting my video which works, then friend sends their video and it blows up. ( the LD_PRELOAD=/usr/lib/libv4lconvert.so.0 is no help )

CODE

writev(7, [{"'\33?\377\21\5)\377\23\f6\37770Z\377SLv\377[T~\377P^r\3777DY\377"..., 133144}, {"", 0}], 2) = 133144
poll([{fd=7, events=POLLIN|POLLOUT}], 1, 4294967295) = 1 ([{fd=7, revents=POLLOUT}])
writev(7, [{"H\2\6\264\215\22\240\3\216\22\240\3\200\2H\0\0\0\230\1\0\30\240\3", 24}, {"\22\20\25\377\21\17\24\377\17\20\24\377\17\20\24\377\17\20\24\377\21\22\26\377\21\22\26\377\20\21\25\377"..., 184320}, {"", 0}], 3) = 184344
munmap(0xe478f000, 1232896)             = 0
poll([{fd=7, events=POLLIN|POLLOUT}], 1, 4294967295) = 1 ([{fd=7, revents=POLLOUT}])
writev(7, [{"<\2\2\0\216\22\240\3\227\4\5\0\217\22\240\3\215\22\240\3!\1\0\0\0\0\0\0\227\7\2\0"..., 220}, {"em\213\377`h\210\377ir\220\377fn\214\377kw\231\377lt\222\377lo\223\377y\177\242\377"..., 19200}, {"", 0}], 3) = 19420
clock_gettime(CLOCK_MONOTONIC, {100989, 286606701}) = 0
poll([{fd=7, events=POLLIN|POLLOUT}], 1, 4294967295) = 1 ([{fd=7, revents=POLLOUT}])
writev(7, [{"<\2\2\0\222\22\240\3\227\10\t\0\3\22\240\3\221\22\240\3\0\0\0\0B\20\240\3\0\0\0\0"..., 800}, {NULL, 0}, {"", 0}], 3) = 800
read(7, 0xb2cef38, 4096)                = -1 EAGAIN (Resource temporarily unavailable)
poll([{fd=3, events=POLLIN}, {fd=8, events=POLLIN}, {fd=7, events=POLLIN}, {fd=73, events=POLLIN}], 4, 96*** glibc detected *** ./skype: malloc(): memory corruption (fast): 0xefc07500 ***
) = 1 ([{fd=3, revents=POLLIN}])
read(3, "A", 1)                         = 1
clock_gettime(CLOCK_MONOTONIC, {100989, 330164322}) = 0
read(7, 0xb2cef38, 4096)                = -1 EAGAIN (Resource temporarily unavailable)
--- SIGSEGV (Segmentation fault) @ 0 (0) ---
madvise(0xffa4d000, 1, MADV_WILLNEED)   = -1 EBADF (Bad file descriptor)
madvise(0xffa4e000, 1, MADV_WILLNEED)   = -1 EBADF (Bad file descriptor)
madvise(0xffa4f000, 1, MADV_WILLNEED)   = -1 EBADF (Bad file descriptor)
madvise(0xffa50000, 1, MADV_WILLNEED)   = -1 EBADF (Bad file descriptor)
madvise(0xffa51000, 1, MADV_WILLNEED)   = -1 ENOMEM (Cannot allocate memory)
rt_sigaction(SIGABRT, {SIG_DFL, [], 0}, NULL, 8) = 0
rt_sigprocmask(SIG_UNBLOCK, [ABRT], NULL, 8) = 0
tgkill(32333, 32333, SIGABRT)           = 0
--- SIGABRT (Aborted) @ 0 (0) ---
+++ killed by SIGABRT +++



The last think I can think of that is different is that I'm running XFS now with inode64 mount option instead of EXT4 on my root and other filesystems.

I've tried skype 2.2.0.35 and 2.1.?? from the recommended skype install repo post in these forums.

Anyway, it works well for messaging and audio only. This is probably more of a Skype issue than anything, but I figured I would start by posting my findings here.
PM
^
mgmartin
 Posted: Sep 15 2011, 06:04 AM
Quote Post


SLF Newbie


Group: Members
Posts: 2
Member No.: 834
Joined: 13-September 11









I found the problem. It seems copying the /usr/share/X11 folder off the partition and back on, and it starts up again. Perhaps something was messed up with the inodes.

I still see this error:

Fontconfig error: "conf.d", line 1: failed reading config file

but Skype starts up. The font error seems to be fixed by moving the /etc/fonts to a non inode64 partition. It seems Skype tries to open the conf.d directory as a file and errors out when the directory is on an inode64 mounted xfs partition.

Error opening directory:
CODE

stat64("/etc/fonts/conf.d", {st_mode=S_IFDIR|0755, st_size=4096, ...}) = 0
open("/etc/fonts/conf.d", O_RDONLY)     = 8
read(8, 0xa782a90, 8192)                = -1 EISDIR (Is a directory)


Compared to the non-inode64 linked directory on a different xfs partition:

CODE

stat64("/etc/fonts/conf.d", {st_mode=S_IFDIR|0755, st_size=4096, ...}) = 0
open("/etc/fonts/conf.d", O_RDONLY|O_NONBLOCK|O_LARGEFILE|O_DIRECTORY|O_CLOEXEC) = 8

PM
^
1 User(s) are reading this topic (1 Guests and 0 Anonymous Users)
0 Members:

Topic Options Reply to this topicStart new topicStart Poll