On 23-Apr-2002 Poul-Henning Kamp wrote: > > This commit detects a memory overwrite problem in the kernel which > happens before we ever get into userland for the first time. > > The commit which causes the problem to appear is my own commit to > subr_disklabel.c (1.65). > > If the block below is put back in subr_disklabel.c the memory overwrite > problem goes away (or at least doesn't happen in GEOM). > > My testbox is a single-cpu machine. > > Something is screwed somewhere...
Uhh, you mean if the dksort_mtx is put back? What if the function doesn't do anything, does it still work? Also, what if you just have the static mtx, maybe the mtx is preventing a buffer overflow from hosing other data? > Poul-Henning > > ] #ifdef notquite > ] /* > ] * Mutex to use when delaying niced I/O bound processes in bioqdisksort(). > ] */ > ] static struct mtx dksort_mtx; > ] static void > ] dksort_init(void) > ] { > ] > ] mtx_init(&dksort_mtx, "dksort", NULL, MTX_DEF); > ] } > ] SYSINIT(dksort, SI_SUB_DRIVERS, SI_ORDER_MIDDLE, dksort_init, NULL) > ] #endif > > > In message <[EMAIL PROTECTED]>, Poul-Henning > Kamp > writes: >>phk 2002/04/23 04:48:45 PDT >> >> Modified files: >> sys/geom geom.h geom_dump.c geom_enc.c >> geom_slice.c geom_subr.c >> Log: >> Introduce some serious paranoia to try to catch a memory overwrite problem >> as early as possible. >> >> Sponsored by: DARPA & NAI Labs >> >> Revision Changes Path >> 1.13 +13 -4 src/sys/geom/geom.h >> 1.7 +1 -0 src/sys/geom/geom_dump.c >> 1.3 +1 -0 src/sys/geom/geom_enc.c >> 1.11 +2 -0 src/sys/geom/geom_slice.c >> 1.8 +46 -2 src/sys/geom/geom_subr.c >> > > -- > Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 > [EMAIL PROTECTED] | TCP/IP since RFC 956 > FreeBSD committer | BSD since 4.3-tahoe > Never attribute to malice what can adequately be explained by incompetence. -- John Baldwin <[EMAIL PROTECTED]> <>< http://www.FreeBSD.org/~jhb/ "Power Users Use the Power to Serve!" - http://www.FreeBSD.org/ To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message