Forced M_NOWAIT in uma_core.c

From: Giorgos Keramidas <keramida_at_freebsd.org>
Date: Sun, 20 Mar 2005 00:08:24 +0200
While stress testing a program with a few thousand threads on a current
system today (FreeBSD 6.0-CURRENT #1: Sat Mar 19 08:40:58 EET 2005), the
following came up on my console:

% malloc(M_WAITOK) of "16", forcing M_NOWAIT with the following non-sleepable locks held:
% exclusive sleep mutex UMA lock r = 0 (0xc0761f20) locked _at_ /usr/src/sys/vm/uma_core.c:1498
% KDB: stack backtrace:
% kdb_backtrace(1,1,1,10,c14454e0) at kdb_backtrace+0x29
% witness_warn(5,0,c06a897b,c0689d86,c051fb39) at witness_warn+0x19a
% uma_zalloc_arg(c14454e0,0,102) at uma_zalloc_arg+0x41
% malloc(10,c06e7320,102,10,c06e7320) at malloc+0xae
% free_unr(c1d60b40,18ae1,c259fcf0,c259a200,c25427e0) at free_unr+0x2f
% thread_fini(c259fcf0,16c) at thread_fini+0x33
% zone_drain(c1445900) at zone_drain+0x170
% zone_foreach(c0618cb0,d46c0cfc,c062a5fa,c069825c,246) at zone_foreach+0x37
% uma_reclaim(c069825c,246,0,d46c0c90,c052168c) at uma_reclaim+0x12
% vm_pageout_scan(0,c07623c0,0,c06aaa6e,5c3) at vm_pageout_scan+0xfe
% vm_pageout(0,d46c0d48,0,c062b3c4,0) at vm_pageout+0x2c3
% fork_exit(c062b3c4,0,d46c0d48) at fork_exit+0xa4
% fork_trampoline() at fork_trampoline+0x8
% --- trap 0x1, eip = 0, esp = 0xd46c0d7c, ebp = 0 ---

Is this a known issue of uma_core.c?
Received on Sun Mar 20 2005 - 03:19:13 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:38:30 UTC