Date: Tue, 13 Jul 2010 16:56:09 -0400 From: Leo Prasath Arulraj <leo@nec-labs.com> To: freebsd-scsi@freebsd.org, mjacob@freebsd.org Subject: Re: "Firmware timed out" with qlogic 2312 FC cards Message-ID: <4C3CD2E9.5050108@nec-labs.com> In-Reply-To: <4C3CC782.7000705@nec-labs.com> References: <4C23655F.1070408@nec-labs.com> <4C3CC782.7000705@nec-labs.com>
index | next in thread | previous in thread | raw e-mail
On 07/13/2010 04:07 PM, Leo Prasath Arulraj wrote: > I am running scsi target mode disk emulation code on a QLogic 2312 FC > card. My OS version is : 9.0-CURRENT FreeBSD . This FreeBSD machine > acts as the target while a separate Linux machine acts as the > initiator accessing the FreeBSD machine as target. > > I am seeing "Firmware timed out" errors on the FreeBSD side : > > > isp0: isp_action code 933 > (targ0:isp0:0:0:0): isp_target_start_ctio: CTIO2[207c] flags 40 xfrlen > 0 offset 4096 > isp0: RISC2HOST ISR 0x2f08013 > isp0: Firmware timed out on command > isp0: data CTIO complete > isp0: isp_handle_platform_ctio: CTIO2[2034] sts 0xb flg 0x840 sns 0 > resid 0 MID > isp0: MIDTERM CTIO[0x2034] done > > > And I see UNDERRUN errors on the Linux side : > > [ 1661.998946] sd 0:0:0:0: [sdb] Unhandled error code > [ 1661.998955] sd 0:0:0:0: [sdb] Result: hostbyte=DID_OK > driverbyte=DRIVER_TIMEOUT > [ 1661.998973] sd 0:0:0:0: [sdb] CDB: Read(10): 28 00 00 00 00 00 00 > 00 08 00 > [ 1661.999024] end_request: I/O error, dev sdb, sector 0 > [ 1662.219095] comp_status 21 scsi_status 0 > [ 1662.219118] scsi(0:0:0) UNDERRUN status detected 0x15-0x0. > resid=0xda fw_resid=0x0 cdb=0x12 os_underflow=0x0 > [ 1662.219137] scsi(0:0:0:0) Dropped frame(s) detected (da of fe > bytes)...firmware reported underrun...retrying command. > > > I changed the ct_timeout field of a CTIO in the isp_freebsd.c file to > 120 from 10: > > isp_freebsd.c:1624: cto->ct_timeout = 120; // was 10 > isp_freebsd.c:1658: cto->ct_timeout = 120 ; // was 10 > > Is there someway I can fix this ? Or should I handle it the userspace > program by retrying commands. I suspect something goes wrong with the > firmware because , after this error happens, subsequents requests also > fail. > > Should I update the firmware assembly file isp_2300.h ? Where should I > get the assembly version of the firmware from ? > > Any help is appreciated. > > Thanks, > Leo. > > > > I also see "PCI DMA Bus Error" on the FreeBSD side : isp0: isp_action code 933 (targ0:isp0:0:0:0): isp_target_start_ctio: CTIO2[1d7c] flags 40 xfrlen 4 offset 0 isp0: RISC2HOST ISR 0x2718013 isp0: PCI DMA Bus Error isp0: data CTIO complete isp0: isp_handle_platform_ctio: CTIO2[1d7c] sts 0x10 flg 0x840 sns 0 resid 0 MID isp0: MIDTERM CTIO[0x1d7c] done This seems to follow the "Firmware timed out" error. Thanks, Leo.help
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?4C3CD2E9.5050108>