kernel/drivers/scsi/lpfc/lpfc.h, branch linux-6.9.y

scsi: lpfc: Use a dedicated lock for ras_fwlog state

2024-03-10T22:56:43Z

To reduce usage of and contention for hbalock, a separate dedicated lock is used to protect ras_fwlog state. Signed-off-by: Justin Tee Link: https://lore.kernel.org/r/20240305200503.57317-8-justintee8345@gmail.com Signed-off-by: Martin K. Petersen

scsi: lpfc: Update lpfc_ramp_down_queue_handler() logic

2024-03-10T22:56:43Z

Typically when an out of resource CQE status is detected, the lpfc_ramp_down_queue_handler() logic is called to help reduce I/O load by reducing an sdev's queue_depth. However, the current lpfc_rampdown_queue_depth() logic does not help reduce queue_depth. num_cmd_success is never updated and is always zero, which means new_queue_depth will always be set to sdev->queue_depth. So, new_queue_depth = sdev->queue_depth - new_queue_depth always sets new_queue_depth to zero. And, scsi_change_queue_depth(sdev, 0) is essentially a no-op. Change the lpfc_ramp_down_queue_handler() logic to set new_queue_depth equal to sdev->queue_depth subtracted from number of times num_rsrc_err was incremented. If num_rsrc_err is >= sdev->queue_depth, then set new_queue_depth equal to 1. Eventually, the frequency of Good_Status frames will signal SCSI upper layer to auto increase the queue_depth back to the driver default of 64 via scsi_handle_queue_ramp_up(). Signed-off-by: Justin Tee Link: https://lore.kernel.org/r/20240305200503.57317-5-justintee8345@gmail.com Signed-off-by: Martin K. Petersen

scsi: lpfc: Copyright updates for 14.4.0.0 patches

2024-02-06T01:51:36Z

Update copyrights to 2024 for files modified in the 14.4.0.0 patch set. Signed-off-by: Justin Tee Link: https://lore.kernel.org/r/20240131185112.149731-18-justintee8345@gmail.com Reviewed-by: Himanshu Madhani Signed-off-by: Martin K. Petersen

scsi: lpfc: Change lpfc_vport load_flag member into a bitmask

2024-02-06T01:51:36Z

In attempt to reduce the amount of unnecessary shost_lock acquisitions in the lpfc driver, change load_flag into an unsigned long bitmask and use clear_bit/test_bit bitwise atomic APIs instead of reliance on shost_lock for synchronization. Also, correct the test for FC_UNLOADING in lpfc_ct_handle_mibreq, which incorrectly tests vport->fc_flag rather than vport->load_flag. Signed-off-by: Justin Tee Link: https://lore.kernel.org/r/20240131185112.149731-16-justintee8345@gmail.com Reviewed-by: Himanshu Madhani Signed-off-by: Martin K. Petersen

scsi: lpfc: Change lpfc_vport fc_flag member into a bitmask

2024-02-06T01:51:36Z

In attempt to reduce the amount of unnecessary shost_lock acquisitions in the lpfc driver, change fc_flag into an unsigned long bitmask and use clear_bit/test_bit bitwise atomic APIs instead of reliance on shost_lock for synchronization. Signed-off-by: Justin Tee Link: https://lore.kernel.org/r/20240131185112.149731-15-justintee8345@gmail.com Reviewed-by: Himanshu Madhani Signed-off-by: Martin K. Petersen

scsi: lpfc: Protect vport fc_nodes list with an explicit spin lock

2024-02-06T01:51:36Z

In attempt to reduce the amount of unnecessary shost_lock acquisitions in the lpfc driver, replace shost_lock with an explicit fc_nodes_list_lock spinlock when accessing vport->fc_nodes lists. Although vport memory region is owned by shost->hostdata, it is driver private memory and an explicit fc_nodes list lock for fc_nodes list mutations is more appropriate than locking the entire shost. Signed-off-by: Justin Tee Link: https://lore.kernel.org/r/20240131185112.149731-14-justintee8345@gmail.com Reviewed-by: Himanshu Madhani Signed-off-by: Martin K. Petersen

scsi: lpfc: Change nlp state statistic counters into atomic_t

2024-02-06T01:51:35Z

There is no reason to use the shost_lock to synchronize an LLDD statistics counter. Convert all the nlp state statistic counters into atomic_t. Corresponding zeroing, increments, and reads are converted to atomic versions. Signed-off-by: Justin Tee Link: https://lore.kernel.org/r/20240131185112.149731-13-justintee8345@gmail.com Reviewed-by: Himanshu Madhani Signed-off-by: Martin K. Petersen

scsi: lpfc: Fix possible file string name overflow when updating firmware

2023-11-15T14:52:57Z

Because file_name and phba->ModelName are both declared a size 80 bytes, the extra ".grp" file extension could cause an overflow into file_name. Define a ELX_FW_NAME_SIZE macro with value 84. 84 incorporates the 4 extra characters from ".grp". file_name is changed to be declared as a char and initialized to zeros i.e. null chars. Signed-off-by: Justin Tee Link: https://lore.kernel.org/r/20231031191224.150862-3-justintee8345@gmail.com Reviewed-by: Himanshu Madhani Signed-off-by: Martin K. Petersen

scsi: lpfc: Do not abuse UUID APIs and LPFC_COMPRESS_VMID_SIZE

2023-08-21T21:13:57Z

The lpfc_vmid_host_uuid is not defined as uuid_t and its usage is not the same as for uuid_t operations (like exporting or importing). Hence replace call to uuid_is_null() by respective memchr_inv() without abusing casting. With that, replace LPFC_COMPRESS_VMID_SIZE with plain number and respective sizeof() to make code robust to changes in the future, if any. Signed-off-by: Andy Shevchenko Link: https://lore.kernel.org/r/20230818155452.875781-1-andriy.shevchenko@linux.intel.com Reviewed-by: Justin Tee Signed-off-by: Martin K. Petersen

scsi: lpfc: Refactor cpu affinity assignment paths

2023-07-23T20:17:07Z

During initialization, a lot of the same logic is used on MSI-X vector CPU affinity assignment. Create a lpfc_next_present_cpu() helper routine, and apply its usage for refactoring purposes. Signed-off-by: Justin Tee Link: https://lore.kernel.org/r/20230712180522.112722-10-justintee8345@gmail.com Signed-off-by: Martin K. Petersen