Date: Mon, 18 Dec 2006 23:22:40 +0100 From: Paolo Pisati <piso@freebsd.org> To: FreeBSD_Current <freebsd-current@freebsd.org>, FreeBSD_Amd64 <amd64@freebsd.org> Subject: FreeBSD/amd64 CURRENT@12/18/06: reproducible hang/panic Message-ID: <20061218222240.GA1255@tin.it>
next in thread | raw e-mail | index | archive | help
Hi, since 17/12/06 i decided to switch my main development box from FreeBSD/i386 to FreeBSD/amd64. My box is composed of: PentiumD 920 (2x2.8Ghz no overclock) 2GB ram Interl 945g make.conf: CPUTYPE?= nocona CFLAGS= -O -pipe KERNCONF= GENERIC With FreeBSD/i386 the box worked fine for months, but with FreeBSD/amd64 i encountered these problems: 1) i'm able to solid hang the box (no panic, no crash, just a "picture" of my desktop) putting a bit of load on it (make -j6 buidworld). 2) sometimes when i rebooted the box, i got a panic/double fault after buffers where synced to disk: unfortunately the double fault msg scrolled up the real panic, so i was unable to see it. While investigating the above problems, i decided to add: options DEBUG_MEMGUARD options DEBUG_REDZONE to my kernel config. Made a buildworld, installed and rebooted. With these 2 options activated, the system panics at boot: MEMGUARD DEBUGGING ALLOCATOR INITIALIZED: MEMGUARD MAP BASE: 0xffffffff80e22000 MEMGUARD MAP LIMIT: 0xffffffff83623000 MEMGUARD MAP SIZE: 41947136 (BYTES) MEMORY MODIFIED AFTER FREE 0xffffff0000039d00 (248) val = 5 @ 0xffffff0000039dd0 kernel trap 9 with interrupts disabled fatal trap 9: general protection fault while in kernel mode cpuid = 0; apic id = 00 inst ptr = ... stack ptr = ... frame ptr = ... code segment = ... proc eflags = resume, iopl = 0 current process = 0 () [thread pid 0 tid 0] stopped at strlen+0x4: cmpb $0, (%rdi) db> bt tracing pid 0 tid 0 td 0xffffffff80940080 strlen() at strlen+0x4 vnsprintf() vnsprintf+0x2e panic() at panic+0x148 mtrash_dtor() at mtrash_dtor uma_zalloc_arg() at uma_zalloc_arg+0x2fc malloc() at malloc+0xe4 init_dynamic_kenv() a init_dynamic_kenv+0x5e mi_startup() at mi_startup+0xc0 btext() at btext+0x2c Now, i dunno if this panic is related to the previous described problems, but i would like to know if i'm the only one to experience such problems. bye, P.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20061218222240.GA1255>