Date: Mon, 20 May 1996 19:43:45 -0700 From: David Greenman <davidg@Root.COM> To: Joe McGuckin <joe@ns.via.net> Cc: hackers@FreeBSD.org, questions@FreeBSD.org Subject: Re: Help - Panic Attacks! Message-ID: <199605210243.TAA11579@Root.COM> In-Reply-To: Your message of "Mon, 20 May 1996 14:33:41 PDT." <199605202133.OAA21060@ns.via.net>
next in thread | previous in thread | raw e-mail | index | archive | help
>Our web machine (FreeBSD 2.1, 128Meg) panicked today - twice in a row.
>
>Can anyone shed some light on this?
>May 19 02:03:56 ovation /kernel: panic: kmem_malloc: kmem_map too small
This is caused by a bug in the calculation of the size of the malloc area,
causing the system to calculate it far too small when a large NMBCLUSTERS
value is used. The attached patch fixes this problem.
-DG
David Greenman
Core-team/Principal Architect, The FreeBSD Project
Index: kern_malloc.c
===================================================================
RCS file: /home/ncvs/src/sys/kern/kern_malloc.c,v
retrieving revision 1.12
retrieving revision 1.12.4.1
diff -c -r1.12 -r1.12.4.1
*** kern_malloc.c 1995/05/30 08:05:33 1.12
--- kern_malloc.c 1996/01/29 11:20:25 1.12.4.1
***************
*** 31,37 ****
* SUCH DAMAGE.
*
* @(#)kern_malloc.c 8.3 (Berkeley) 1/4/94
! * $Id: kern_malloc.c,v 1.12 1995/05/30 08:05:33 rgrimes Exp $
*/
#include <sys/param.h>
--- 31,37 ----
* SUCH DAMAGE.
*
* @(#)kern_malloc.c 8.3 (Berkeley) 1/4/94
! * $Id: kern_malloc.c,v 1.12.4.1 1996/01/29 11:20:25 davidg Exp $
*/
#include <sys/param.h>
***************
*** 39,44 ****
--- 39,45 ----
#include <sys/proc.h>
#include <sys/kernel.h>
#include <sys/malloc.h>
+ #include <sys/mbuf.h>
#include <vm/vm.h>
#include <vm/vm_kern.h>
***************
*** 371,384 ****
#if (MAXALLOCSAVE < CLBYTES)
ERROR!_kmeminit:_MAXALLOCSAVE_too_small
#endif
! npg = VM_KMEM_SIZE/ NBPG;
! if( npg > cnt.v_page_count)
! npg = cnt.v_page_count;
kmemusage = (struct kmemusage *) kmem_alloc(kernel_map,
(vm_size_t)(npg * sizeof(struct kmemusage)));
kmem_map = kmem_suballoc(kernel_map, (vm_offset_t *)&kmembase,
! (vm_offset_t *)&kmemlimit, (vm_size_t)(npg * NBPG), FALSE);
#ifdef KMEMSTATS
for (indx = 0; indx < MINBUCKET + 16; indx++) {
if (1 << indx >= CLBYTES)
--- 372,384 ----
#if (MAXALLOCSAVE < CLBYTES)
ERROR!_kmeminit:_MAXALLOCSAVE_too_small
#endif
! npg = (nmbclusters * MCLBYTES + VM_KMEM_SIZE) / PAGE_SIZE;
kmemusage = (struct kmemusage *) kmem_alloc(kernel_map,
(vm_size_t)(npg * sizeof(struct kmemusage)));
kmem_map = kmem_suballoc(kernel_map, (vm_offset_t *)&kmembase,
! (vm_offset_t *)&kmemlimit, (vm_size_t)(npg * PAGE_SIZE),
! FALSE);
#ifdef KMEMSTATS
for (indx = 0; indx < MINBUCKET + 16; indx++) {
if (1 << indx >= CLBYTES)
***************
*** 387,393 ****
bucket[indx].kb_elmpercl = CLBYTES / (1 << indx);
bucket[indx].kb_highwat = 5 * bucket[indx].kb_elmpercl;
}
! for (indx = 0; indx < M_LAST; indx++)
! kmemstats[indx].ks_limit = npg * NBPG * 6 / 10;
#endif
}
--- 387,399 ----
bucket[indx].kb_elmpercl = CLBYTES / (1 << indx);
bucket[indx].kb_highwat = 5 * bucket[indx].kb_elmpercl;
}
! /*
! * Limit maximum memory for each type to 60% of malloc area size or
! * 60% of physical memory, whichever is smaller.
! */
! for (indx = 0; indx < M_LAST; indx++) {
! kmemstats[indx].ks_limit = min(cnt.v_page_count * PAGE_SIZE,
! (npg * PAGE_SIZE - nmbclusters * MCLBYTES)) * 6 / 10;
! }
#endif
}
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199605210243.TAA11579>