HELP! iSCSI qla4010 HBA driver problems.

Jimmy McMillan jimmy at freshstation.org
Tue May 31 19:13:38 UTC 2005


Ok, scratch that, it still isn't working...Still getting the mailbox 
errors...  Cannot even soft reboot it.  I have to pull the power from 
the machine because it's stuck trying to talk with the SAN.  More info 
if needed.  Thanks!

scsi1: qla4xxx_mailbox_command: Trying to execute a mailbox request, 
while another one is interrupting
Service existing interrupt first
scsi1: qla4xxx_mailbox_command: **** FAILED, cmd = 00000069, sts = 
00000000 ****
0x00 mailbox[0]      = 0x00004000
0x04 mailbox[1]      = 0x00000001
0x08 mailbox[2]      = 0x00040002
0x0C mailbox[3]      = 0x00000010
0x10 mailbox[4]      = 0x00000001
0x14 mailbox[5]      = 0x00000000
0x18 mailbox[6]      = 0x00000000
0x1C mailbox[7]      = 0xFFFFAD37
0x20 flash_address   = 0x00057DFF
0x24 flash_data      = 0x000000FF
0x28 ctrl_status     = 0x00000040
0x2C nvram           = 0x00000008
0x38 req_q_in        = 0x00000000
0x3C rsp_q_out       = 0x00000000
0x50 ext_hw_conf     = 0x00001916
0x58 port_ctrl       = 0x00000C1F
0x5C port_status     = 0x0001C61D
0x80 req_q_out       = 0x00000000
0xE0 gp_out          = 0x00000003
0xE4 gp_in           = 0x0000000E
0xFC port_err_status = 0x00000000
scsi1: qla4extioctl_query_hba_iscsi_portal: MBOX_CMD_GET_FW_STATE failed 
w/ status 0000

Jimmy McMillan wrote:

> Awesome, got it compiled, and probed.  So far so good.  I got a few 
> warnings included at compile time, but all in all it's working.  I'm 
> gonna setup random writes to the device for a day to see if it 
> eventually fails.  It's so random.  Thanks alot Andy!
>
>
> qla4xxx -- Building the qla4xxx driver...
> make: Entering directory `/lib/modules/2.6.11-1.27_FC3smp/build'
>  CC [M]  /usr/src/qlogic/v5.00.02-1/5.00.02/ql4_os.o
>  CC [M]  /usr/src/qlogic/v5.00.02-1/5.00.02/ql4_init.o
>  CC [M]  /usr/src/qlogic/v5.00.02-1/5.00.02/ql4_mbx.o
>  CC [M]  /usr/src/qlogic/v5.00.02-1/5.00.02/ql4_iocb.o
>  CC [M]  /usr/src/qlogic/v5.00.02-1/5.00.02/ql4_isr.o
> /usr/src/qlogic/v5.00.02-1/5.00.02/ql4_isr.c: In function 
> `qla4xxx_check_and_copy_sense':
> /usr/src/qlogic/v5.00.02-1/5.00.02/ql4_isr.c:161: warning: ISO C90 
> forbids mixed declarations and code
> /usr/src/qlogic/v5.00.02-1/5.00.02/ql4_os.c: In function 
> `qla4xxx_config_dma_addressing':
> /usr/src/qlogic/v5.00.02-1/5.00.02/ql4_os.c:876: warning: ignoring 
> return value of `pci_set_consistent_dma_mask', declared with attribute 
> warn_unused_result
> /usr/src/qlogic/v5.00.02-1/5.00.02/ql4_os.c:884: warning: ignoring 
> return value of `pci_set_dma_mask', declared with attribute 
> warn_unused_result
>  CC [M]  /usr/src/qlogic/v5.00.02-1/5.00.02/ql4_isns.o
>  CC [M]  /usr/src/qlogic/v5.00.02-1/5.00.02/ql4_nvram.o
>  CC [M]  /usr/src/qlogic/v5.00.02-1/5.00.02/ql4_dbg.o
>  CC [M]  /usr/src/qlogic/v5.00.02-1/5.00.02/ql4_cfg.o
>  CC [M]  /usr/src/qlogic/v5.00.02-1/5.00.02/ql4_cfgln.o
>  CC [M]  /usr/src/qlogic/v5.00.02-1/5.00.02/ql4_fo.o
>  CC [M]  /usr/src/qlogic/v5.00.02-1/5.00.02/ql4_foio.o
>  CC [M]  /usr/src/qlogic/v5.00.02-1/5.00.02/ql4_foioctl.o
>  CC [M]  /usr/src/qlogic/v5.00.02-1/5.00.02/ql4_foln.o
>  CC [M]  /usr/src/qlogic/v5.00.02-1/5.00.02/ql4_xioct.o
>  CC [M]  /usr/src/qlogic/v5.00.02-1/5.00.02/ql4_inioct.o
>  CC [M]  /usr/src/qlogic/v5.00.02-1/5.00.02/ql4_32ioctl.o
>  LD [M]  /usr/src/qlogic/v5.00.02-1/5.00.02/qla4xxx.o
>  Building modules, stage 2.
>  MODPOST
>  CC      /usr/src/qlogic/v5.00.02-1/5.00.02/qla4xxx.mod.o
>  LD [M]  /usr/src/qlogic/v5.00.02-1/5.00.02/qla4xxx.ko
> make: Leaving directory `/lib/modules/2.6.11-1.27_FC3smp/build'
>
>
>
>
> Andy Green wrote:
>
>> -----BEGIN PGP SIGNED MESSAGE-----
>> Hash: SHA1
>>
>> Jimmy McMillan wrote:
>> | Also here's a build error I get when compiling the module.
>>
>> | /usr/src/qlogic/v5.00.02-1/5.00.02/ql4_os.c: In function
>> | `qla4xxx_probe_adapter':
>> | /usr/src/qlogic/v5.00.02-1/5.00.02/ql4_os.c:195: sorry, unimplemented:
>> | inlining failed in call to 'qla4xxx_config_dma_addressing': function
>> | body not available
>>
>> These are likely genuine problems with the source which the later
>> version of the compiler you have now is detecting and complaining about,
>> whereas previous versions of the compiler silently accepted the error.
>>
>> Remove the "inline" token at the function declarations and it should
>> continue.
>>
>> - -Andy
>> -----BEGIN PGP SIGNATURE-----
>> Version: GnuPG v1.4.1 (GNU/Linux)
>> Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org
>>
>> iD8DBQFCnHs8jKeDCxMJCTIRAhFzAJ9EDEffZ3ihzpaFkitYpzMxMDX0FQCfaviy
>> NfsjZg1If4j35NU7qm1Buw0=
>> =XJN4
>> -----END PGP SIGNATURE-----
>>
>




More information about the users mailing list