We have a News server running on Compaq Proliant 3000 hardware, with NCR/Symbios 875 based SCSI controllers. This machine ran extremely stable (but using only one processor) on FreeBSD-2.2.8. Due to the News server, this machine sees reasonably heavy disk and network traffic. On FreeBSD 3.x we have been able to use both processors, but the system has never been stable. We've typically had crashes about once a week. We have suspected the problem described in the recent "PCI DMA lockups in 3.2 (3.3 maybe?)" thread, > 1) Random crashes in FXP interrupt or low-level IP code. Something is > clobbering the kernel stack--possibly the NCR driver, since using an > Adaptec made the problem stop, as did a backport of the CAM driver > Peter Wemm tried. This was on an N440BX, which is becoming quite > common in server applications. Other installations are apparantly > seeing the same problem on this hardware. but have never been able to verify this. Below are included parts of some kernel backtraces to show where it crashes. Anyway, we recently switched to the SYM driver (version 0.12.0). The machine has now been up for 16 days without a single crash, and we're very optimistic that the SYM driver has cured the problem. Kudos and thanks to Gerard Roudier! Steinar Haug, Nethelp consulting, [EMAIL PROTECTED] ---------------------------------------------------------------------- crash 0,2,5: #1 0xc0144299 in panic #2 0xc015ac7e in m_copym #3 0xc019286a in tcp_output #4 0xc0194106 in tcp_usr_send #5 0xc015c4b2 in sosend #6 0xc01525dc in soo_write #7 0xc014f46a in dofilewrite #8 0xc014f373 in write #9 0xc021f39b in syscall crash 1,3,4: #1 0xc0144299 in panic #2 0xc015dfe0 in sbdrop #3 0xc015df69 in sbflush #4 0xc0194707 in tcp_disconnect #5 0xc0193ee6 in tcp_usr_disconnect #6 0xc015bf68 in sodisconnect #7 0xc015bd9e in soclose #8 0xc01527ab in soo_close #9 0xc013cf1c in closef #10 0xc013c56d in close #11 0xc021f39b in syscall crash 6 (starts with the same write/dofilewrite sequence as 0,2,5): #1 0xc0144299 in panic #2 0xc015ac7e in m_copym #3 0xc019286a in tcp_output #4 0xc0191ae8 in tcp_input #5 0xc018d07c in ip_input #6 0xc018d0db in ipintr #7 0xc020fe85 in swi_net_next () #8 0xc0226b94 in splx #9 0xc021d86f in pmap_clear_modify #10 0xc01f8775 in vm_page_set_validclean #11 0xc01642a0 in vfs_page_set_valid #12 0xc0164423 in vfs_clean_pages #13 0xc0161fbe in bdwrite #14 0xc0165ce0 in cluster_write #15 0xc01e656b in ffs_write #16 0xc016f572 in vn_write #17 0xc014f46a in dofilewrite #18 0xc014f373 in write #19 0xc021f39b in syscall crash 7 (similar to 0,2,5 followed by reboot and a new crash): #1 0xc0144299 in panic #2 0xc021f120 in trap_fatal #3 0xc021ed9b in trap_pfault #4 0xc021e9fa in trap #5 0xc01cffcc in fxp_add_rfabuf #6 0xc01cf410 in fxp_intr #7 0xc0137261 in intr_mux #8 0xc0226b94 in splx #9 0xc0163233 in getblk #10 0xc01e588e in ffs_sbupdate #11 0xc01e5494 in ffs_sync #12 0xc016b4e7 in sync #13 0xc0143e11 in boot #14 0xc0144299 in panic #15 0xc021f120 in trap_fatal #16 0xc021ed9b in trap_pfault #17 0xc021e9fa in trap #18 0xc015c362 in sosend #19 0xc01525dc in soo_write #20 0xc014f46a in dofilewrite #21 0xc014f373 in write #22 0xc021f39b in syscall crash 8 (similar to 0,2,5 followed by reboot and a new crash): #1 0xc0144299 in panic #2 0xc021f120 in trap_fatal #3 0xc021ed9b in trap_pfault #4 0xc021e9fa in trap #5 0xc01cffcc in fxp_add_rfabuf #6 0xc01cf410 in fxp_intr #7 0xc0137261 in intr_mux #8 0xc0226b94 in splx #9 0xc0163233 in getblk #10 0xc0161bad in bread #11 0xc01dd924 in ffs_update #12 0xc01e6fae in ffs_fsync #13 0xc01e5342 in ffs_sync #14 0xc016b4e7 in sync #15 0xc0143e11 in boot #16 0xc0144299 in panic #17 0xc021f120 in trap_fatal #18 0xc021ed9b in trap_pfault #19 0xc021e9fa in trap #20 0xc015c362 in sosend #21 0xc01525dc in soo_write #22 0xc014f46a in dofilewrite #23 0xc014f373 in write #24 0xc021f39b in syscall To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
