linux/include/scsi, branch v2.6.13

[SCSI] fix target scanning oops with fc transport class

2005-08-08T22:14:55Z

We have some nasty issues with 2.6.12-rc6. Any request to scan on the lpfc or qla2xxx FC adapters will oops. What is happening is the system is defaulting to non-transport registered targets, which inherit the parent of the scan. On this second scan, performed by the attribute, the parent becomes the shost instead of the rport. The slave functions in the 2 FC adapters use starget_to_rport() routines, which incorrectly map the shost as an rport pointer. Additionally, this pointed out other weaknesses: - If the target structure is torn down outside of the transport, we have no method for it to be regenerated at the proper parent. - We have race conditions on the target being allocated by both the midlayer scan (parent=shost) and by the fc transport (parent=rport). Signed-off-by: James Bottomley

[SCSI] fix function prototype warning

2005-07-14T16:54:17Z

int_to_scsilun() takes a pointer to a struct scsi_lun in it's prototype, so add this structure to scsi_device.h to avoid declaration inside function prototype warnings. Signed-off-by: James Bottomley

[SCSI] add int_to_scsilun() function

2005-07-14T15:21:27Z

One of the issues we had was reverting the midlayers lun value into the 8byte lun value that we wanted to send to the device. Historically, there's been some combination of byte swapping, setting high/low, etc. There's also been no common thread between how our driver did it and others. I also got very confused as to why byteswap routines were being used. Anyway, this patch is a LLDD-callable function that reverts the midlayer's lun value, stored in an int, to the 8-byte quantity (note: this is not the real 8byte quantity, just the same amount that scsilun_to_int() was able to convert and store originally). This also solves the dilemma of the thread: http://marc.theaimsgroup.com/?l=linux-kernel&m=112116767118981&w=2 A patch for the lpfc driver to use this function will be along in a few days (batched with other patches). Signed-off-by: James Bottomley

[SCSI] add TYPE_RBC to our type table

2005-07-11T17:38:33Z

Here's a tiny update that means we print the correct ASCII type information Signed-off-by: James Bottomley

[SCSI] remove scsi_cmnd->state

2005-06-26T17:16:24Z

We never look at it except for the old megaraid driver that abuses it for sending internal commands. That usage can be fixed easily because those internal commands are single-threaded by a mutex and we can easily use a completion there. Signed-off-by: James Bottomley

[SCSI] remove scsi_cmnd->owner

2005-06-26T17:15:28Z

never checked anywhere Signed-off-by: James Bottomley

[SCSI] remove scsi_cmnd->abort_reason

2005-06-26T17:14:46Z

Never used for anything but printing it out in debug routines. Signed-off-by: James Bottomley

[SCSI] remove scsi_cmnd.eh_state

2005-06-26T17:14:10Z

it's never set to anything, and just three broken drivers are looking at it and doing odd things. Signed-off-by: James Bottomley

[SCSI] remove scsi_set_device

2005-06-26T13:31:47Z

scsi_add_host is the proper place to set the device, but people copy the scsi_set_device usage from older drivers again and again. note that this leaves some legacy drivers like qlogicisp/qlogicfc without pci association in sysfs, but they're scheduled to go away soon anyway. Signed-off-by: James Bottomley

[PATCH] I2O: Adaptec specific SG_IO access, firmware access through sysfs and 2400A workaround

2005-06-24T07:05:28Z

Changes: - Provide SG_IO access to BLOCK and EXECUTIVE class on Adaptec controllers - Use PRIVATE messages in SCSI-OSM because on some controllers normal SCSI class commands like READ or READ CAPACITY cause errors - Use new DMA and SG list creation function - Added workaround to limit sectors per request for Adaptec 2400A controllers Signed-off-by: Markus Lidel Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds