Date: Fri, 30 Aug 2002 13:19:38 +0200 (CEST) From: Pawel Jakub Dawidek <nick@garage.freebsd.pl> To: FreeBSD-gnats-submit@FreeBSD.org Subject: kern/42198: Kernel panics or system hangs up with big SC_HISTORY_SIZE and VESA. Message-ID: <20020830111938.597EB3ABD40@milla.ask33.net>
next in thread | raw e-mail | index | archive | help
>Number: 42198 >Category: kern >Synopsis: Kernel panics or system hangs up with big SC_HISTORY_SIZE and VESA. >Confidential: no >Severity: critical >Priority: high >Responsible: freebsd-bugs >State: open >Quarter: >Keywords: >Date-Required: >Class: sw-bug >Submitter-Id: current-users >Arrival-Date: Fri Aug 30 04:20:02 PDT 2002 >Closed-Date: >Last-Modified: >Originator: Pawel Jakub Dawidek >Release: FreeBSD 4.6-STABLE i386 >Organization: Private >Environment: System: FreeBSD milla.ask33.net 4.6-STABLE FreeBSD 4.6-STABLE #9: Wed Jul 31 21:18:46 CEST 2002 root@milla.ask33.net:/usr/obj/usr/src/sys/MILLA i386 >Description: There are some serious problems with big values of SC_HISTORY_SIZE and VESA video driver. Kernel panics when we are trying to load vesa.ko module and SC_HISTORY_SIZE is big or system hangs up when SC_HISTORY_SIZE is big and VESA is compiled in kernel. >How-To-Repeat: There are two ways: 1. Compile kernel with big (for example 5000) SC_HISTORY_SZIE and without VESA. Then reboot machine and do # kldload vesa Kernel sould panic. 2. Compile kernel with big (for example 5000) SC_HISTORY_SIZE and with VESA support, then reboot machine and: # vidcontrol 132x60 then: # cat <some_big_file> few times. When history buffer fills up system will hangs up. But there is a chance that those are two diffrent bugs. >Fix: I don't know exactly. Backtrace after panic shows lncintr_sc() function from /sys/i386/isa/if_lnc.c. But I don't think secound problem (with system hangs up) is there. Problem could be here (from /sys/dev/syscons/schistory.c): #define SC_MAX_HISTORY_SIZE (1000 * MAXCONS * NSC) NSC is as I suppose number of chars per line. But where it is defined? I can't grep it. And there is a chance that this is to less for 132 chars per line, because maybe somewhere buffer size is counted as number_of_lines * x_console_size and we got overflow. But I could be wrong:) >Release-Note: >Audit-Trail: >Unformatted: To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-bugs" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20020830111938.597EB3ABD40>