* 2.4.0-prerelease shmget woes. @ 2001-01-01 19:06 Gregory McLean 2001-01-01 19:34 ` Gregory McLean 0 siblings, 1 reply; 4+ messages in thread From: Gregory McLean @ 2001-01-01 19:06 UTC (permalink / raw) To: linux-kernel Hi, I just compiled and booted up 2.4.0-prerelease, dropped into X-Windows and now gtk/gnome apps are giving the following warning: Gdk-WARNING **: shmget failed! Any ideas on what changed in the shm space to cause this? This is a PII 333 RH 7.0 system 2.4.0-test10 worked just fine on this machine. /dev/shm is mounted: none 0 0 0 - /dev/shm The /usr/src/linux/Documentation/Changes file only mentions that this need be mounted. Tips, clues welcome. -- Greg - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org Please read the FAQ at http://www.tux.org/lkml/ ^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: 2.4.0-prerelease shmget woes. 2001-01-01 19:06 2.4.0-prerelease shmget woes Gregory McLean @ 2001-01-01 19:34 ` Gregory McLean 2001-01-02 21:38 ` Christoph Rohland 0 siblings, 1 reply; 4+ messages in thread From: Gregory McLean @ 2001-01-01 19:34 UTC (permalink / raw) To: linux-kernel On Mon, 1 Jan 2001, Gregory McLean wrote: > > Hi, I just compiled and booted up 2.4.0-prerelease, dropped into X-Windows > and now gtk/gnome apps are giving the following warning: > > Gdk-WARNING **: shmget failed! > > Any ideas on what changed in the shm space to cause this? > > This is a PII 333 RH 7.0 system 2.4.0-test10 worked just fine on this > machine. Per request here is an strace around where the shmget fails: (xmms is the target app running Full file available if needed) getrlimit(0x3, 0xbffff430, 0xbffff4f8, 0xbffff430, 0xbffff430) = 0 setrlimit(RLIMIT_STACK, {rlim_cur=2044*1024, rlim_max=RLIM_INFINITY}) = 0 brk(0x8172000) = 0x8172000 pipe([5, 6]) = 0 clone(child_stack=0x8171ac8, flags=CLONE_VM|CLONE_FS|CLONE_FILES|CLONE_SIGHAND) = 6299 write(6, "\0\0\0\0\5\0\0\0\0\0\0\0\0\200\31@\0\0\0\0\30l\1@\360A"..., 148) = 148 rt_sigprocmask(SIG_SETMASK, NULL, [RT_0], 8) = 0 write(6, "\0\17\35@\0\0\0\0\0\0\0\0 \223\6\10\0\0\0\0\0\0\0\200\0"..., 148) = 14 8 rt_sigprocmask(SIG_SETMASK, NULL, [RT_0], 8) = 0 rt_sigsuspend([] <unfinished ...> --- SIGRT_0 (Real-time signal 0) --- <... rt_sigsuspend resumed> ) = -1 EINTR (Interrupted system call) sigreturn() = ? (mask now []) shmget(IPC_PRIVATE, 32768, IPC_CREAT|0x1ff|0666) = -1 ENOSPC (No space left on d evice) write(2, "\n", 1 ) = 1 write(2, "Gdk", 3Gdk) = 3 write(2, "-", 1-) = 1 write(2, "WARNING **: ", 12WARNING **: ) = 12 write(2, "shmget failed!", 14shmget failed!) = 14 write(2, "\n", 1 ) = 1 brk(0x817b000) = 0x817b000 brk(0x8184000) = 0x8184000 brk(0x818d000) = 0x818d000 brk(0x8196000) = 0x8196000 brk(0x819f000) = 0x819f000 brk(0x81a8000) = 0x81a8000 brk(0x81b9000) = 0x81b9000 write(3, "\211\30\2\0\2\0\0\0\20\0\10\0\30\0op_WIN_SUPPORTING_"..., 40) = 40 read(3, "\1`)\0\0\0\0\0\34\1\0\0\0\0\0\0\0\0\0\0\1\0\0\0\200\252"..., 32) = 32 write(3, "\20\0\5\0\n\0\0\0_WIN_LAYERIN", 20) = 20 read(3, "\1\0*\0\0\0\0\0#\1\0\0\0\0\0\0\1\0\0\0\1\0\0\0\200\252"..., 32) = 32 write(3, "\20\0\5\0\n\0\0\0_WIN_HINTSIN", 20) = 20 (bunch more garbage of xmms talking to the x server) >From what I can tell its saying no space left on device. >From proc: --------- cat /proc/sys/kernel/shmmax 33554432 cat /proc/sys/kernel/shmall 0 cat /proc/sys/kernel/shmmni 4096 - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org Please read the FAQ at http://www.tux.org/lkml/ ^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: 2.4.0-prerelease shmget woes. 2001-01-01 19:34 ` Gregory McLean @ 2001-01-02 21:38 ` Christoph Rohland 2001-01-05 18:53 ` Gregory McLean 0 siblings, 1 reply; 4+ messages in thread From: Christoph Rohland @ 2001-01-02 21:38 UTC (permalink / raw) To: Gregory McLean; +Cc: linux-kernel Gregory McLean <gregm@comstar.net> writes: > cat /proc/sys/kernel/shmall > 0 That's your problem. Your limit for overall shm pages is zero. So you cannot allocate any shm segments. echo 2000000 > /proc/sys/kernel/shmall and check /etc/sysctl.conf or wherever your system stores kernel parameters to restore on boot. Greetings Christoph - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org Please read the FAQ at http://www.tux.org/lkml/ ^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: 2.4.0-prerelease shmget woes. 2001-01-02 21:38 ` Christoph Rohland @ 2001-01-05 18:53 ` Gregory McLean 0 siblings, 0 replies; 4+ messages in thread From: Gregory McLean @ 2001-01-05 18:53 UTC (permalink / raw) To: Christoph Rohland; +Cc: linux-kernel On 2 Jan 2001, Christoph Rohland wrote: > Gregory McLean <gregm@comstar.net> writes: > > cat /proc/sys/kernel/shmall > > 0 > > That's your problem. Your limit for overall shm pages is zero. So you > cannot allocate any shm segments. > > echo 2000000 > /proc/sys/kernel/shmall > > and check /etc/sysctl.conf or wherever your system stores kernel > parameters to restore on boot. Yeah for some odd reason RedHack stuffed a 0 in there *grumble* That fixed it. > > Greetings > Christoph > - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org Please read the FAQ at http://www.tux.org/lkml/ ^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2001-01-05 18:54 UTC | newest] Thread overview: 4+ messages (download: mbox.gz follow: Atom feed -- links below jump to the message on this page -- 2001-01-01 19:06 2.4.0-prerelease shmget woes Gregory McLean 2001-01-01 19:34 ` Gregory McLean 2001-01-02 21:38 ` Christoph Rohland 2001-01-05 18:53 ` Gregory McLean
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox