Date: Sat, 22 Jan 2011 16:07:04 +0200 From: Nikos Vassiliadis <nvass9573@gmx.com> To: freebsd-geom@freebsd.org Subject: kmem_map panic with gjournal Message-ID: <4D3AE488.7050609@gmx.com>
next in thread | raw e-mail | index | archive | help
Hi, I am seeing a "kmem_map too small" panic on a low resources machine (384MB of RAM). After lowering kern.geom.journal.cache.limit the panic is gone, yet I am not sure that this is the "best" solution. Any thoughts? Backtrace from RELENG_8. The panic also occurs on HEAD. Unread portion of the kernel message buffer: panic: kmem_malloc(131072): kmem_map too small: 111280128 total allocated cpuid = 0 KDB: stack backtrace: db_trace_self_wrapper(c0d3b1ec,d1be5568,c090107c,c31d1870,d1be5568,...) at db_trace_self_wrapper+0x26 kdb_backtrace(c0d68c13,0,c0d5c9b8,d1be55d8,0,...) at kdb_backtrace+0x2a panic(c0d5c9b8,20000,6a20000,c0d5c9b2,7d0,...) at panic+0x117 kmem_malloc(c199008c,20000,2,d1be5640,c0b7cc90,...) at kmem_malloc+0x28a page_alloc(0,20000,d1be5633,2,2c41a72,...) at page_alloc+0x27 uma_large_malloc(20000,2,d1be56a8,2,20000,...) at uma_large_malloc+0x50 malloc(20000,c10aa080,2,111,c9415000,...) at malloc+0x80 gj_malloc(c93f5000,ca3ed000,20000,42788000,2,...) at gj_malloc+0x9b g_journal_new_bio(427a8000,2,10ee800,0,c9415000,...) at g_journal_new_bio+0x62 g_journal_insert(427a8000,2,10ee800,0,c9415000,...) at g_journal_insert+0xe0 g_journal_flush(c31ae610,0,c10a8b26,87e,7d0,...) at g_journal_flush+0x36a g_journal_worker(c31ae600,d1be5d28,0,0,0,...) at g_journal_worker+0x162d fork_exit(c10a64e0,c31ae600,d1be5d28) at fork_exit+0x91 fork_trampoline() at fork_trampoline+0x8 --- trap 0, eip = 0, esp = 0xd1be5d60, ebp = 0 --- Uptime: 7m11s Physical memory: 367 MB Dumping 177 MB: 162 146 130 114 98 82 66 50 34 18 2 Thanks, Nikos
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?4D3AE488.7050609>