| Message ID | 20220610232915.2916712-1-bvanassche@acm.org (mailing list archive) |
|---|---|
| State | Superseded |
| Headers | show |
| Series | scsi: ufs: Fix a race between the interrupt handler and the reset handler | expand |
On 11/06/22 02:29, Bart Van Assche wrote: > Prevent that both the interrupt handler and the reset handler try to > complete a request at the same time. This patch is the result of the > analysis of the following crash: > > Unable to handle kernel NULL pointer dereference at virtual address 0000000000000120 > CPU: 0 PID: 0 Comm: swapper/0 Tainted: G OE 5.10.107-android13-4-00051-g1e48e8970cca-ab8664745 #1 > pc : ufshcd_release_scsi_cmd+0x30/0x46c > lr : __ufshcd_transfer_req_compl+0x4fc/0x9c0 > Call trace: > ufshcd_release_scsi_cmd+0x30/0x46c > __ufshcd_transfer_req_compl+0x4fc/0x9c0 > ufshcd_poll+0xf0/0x208 > ufshcd_sl_intr+0xb8/0xf0 > ufshcd_intr+0x168/0x2f4 > __handle_irq_event_percpu+0xa0/0x30c > handle_irq_event+0x84/0x178 > handle_fasteoi_irq+0x150/0x2e8 > __handle_domain_irq+0x114/0x1e4 > gic_handle_irq.31846+0x58/0x300 > el1_irq+0xe4/0x1c0 > cpuidle_enter_state+0x3ac/0x8c4 > do_idle+0x2fc/0x55c > cpu_startup_entry+0x84/0x90 > kernel_init+0x0/0x310 > start_kernel+0x0/0x608 > start_kernel+0x4ec/0x608 > > Signed-off-by: Bart Van Assche <bvanassche@acm.org> > --- > drivers/scsi/ufs/ufshcd.c | 20 +++++++++++++------- > 1 file changed, 13 insertions(+), 7 deletions(-) > > diff --git a/drivers/scsi/ufs/ufshcd.c b/drivers/scsi/ufs/ufshcd.c > index 1fb3a8b9b03e..279691ff3562 100644 > --- a/drivers/scsi/ufs/ufshcd.c > +++ b/drivers/scsi/ufs/ufshcd.c > @@ -6966,6 +6966,7 @@ int ufshcd_exec_raw_upiu_cmd(struct ufs_hba *hba, > */ > static int ufshcd_eh_device_reset_handler(struct scsi_cmnd *cmd) > { > + unsigned long flags, completed_reqs = 0; > struct Scsi_Host *host; > struct ufs_hba *hba; > u32 pos; > @@ -6984,13 +6985,18 @@ static int ufshcd_eh_device_reset_handler(struct scsi_cmnd *cmd) > } > > /* clear the commands that were pending for corresponding LUN */ > - for_each_set_bit(pos, &hba->outstanding_reqs, hba->nutrs) { > - if (hba->lrb[pos].lun == lun) { > - err = ufshcd_clear_cmd(hba, pos); > - if (err) > - break; > - __ufshcd_transfer_req_compl(hba, 1U << pos); > - } > + spin_lock_irqsave(&hba->outstanding_lock, flags); > + for_each_set_bit(pos, &hba->outstanding_reqs, hba->nutrs) > + if (hba->lrb[pos].lun == lun) > + __set_bit(pos, &completed_reqs); > + hba->outstanding_reqs &= ~completed_reqs; > + spin_unlock_irqrestore(&hba->outstanding_lock, flags); > + > + for_each_set_bit(pos, &completed_reqs, hba->nutrs) { > + err = ufshcd_clear_cmd(hba, pos); > + if (err) > + break; Having cleared the bit in hba->outstanding_reqs, shouldn't we always complete the request? i.e. we should not 'break' here > + __ufshcd_transfer_req_compl(hba, 1U << pos); > } > > out:
On 6/12/22 23:28, Adrian Hunter wrote: > Having cleared the bit in hba->outstanding_reqs, shouldn't we > always complete the request? i.e. we should not 'break' here Agreed. I will fix this, retest and repost. Thanks, Bart.
diff --git a/drivers/scsi/ufs/ufshcd.c b/drivers/scsi/ufs/ufshcd.c index 1fb3a8b9b03e..279691ff3562 100644 --- a/drivers/scsi/ufs/ufshcd.c +++ b/drivers/scsi/ufs/ufshcd.c @@ -6966,6 +6966,7 @@ int ufshcd_exec_raw_upiu_cmd(struct ufs_hba *hba, */ static int ufshcd_eh_device_reset_handler(struct scsi_cmnd *cmd) { + unsigned long flags, completed_reqs = 0; struct Scsi_Host *host; struct ufs_hba *hba; u32 pos; @@ -6984,13 +6985,18 @@ static int ufshcd_eh_device_reset_handler(struct scsi_cmnd *cmd) } /* clear the commands that were pending for corresponding LUN */ - for_each_set_bit(pos, &hba->outstanding_reqs, hba->nutrs) { - if (hba->lrb[pos].lun == lun) { - err = ufshcd_clear_cmd(hba, pos); - if (err) - break; - __ufshcd_transfer_req_compl(hba, 1U << pos); - } + spin_lock_irqsave(&hba->outstanding_lock, flags); + for_each_set_bit(pos, &hba->outstanding_reqs, hba->nutrs) + if (hba->lrb[pos].lun == lun) + __set_bit(pos, &completed_reqs); + hba->outstanding_reqs &= ~completed_reqs; + spin_unlock_irqrestore(&hba->outstanding_lock, flags); + + for_each_set_bit(pos, &completed_reqs, hba->nutrs) { + err = ufshcd_clear_cmd(hba, pos); + if (err) + break; + __ufshcd_transfer_req_compl(hba, 1U << pos); } out:
Prevent that both the interrupt handler and the reset handler try to complete a request at the same time. This patch is the result of the analysis of the following crash: Unable to handle kernel NULL pointer dereference at virtual address 0000000000000120 CPU: 0 PID: 0 Comm: swapper/0 Tainted: G OE 5.10.107-android13-4-00051-g1e48e8970cca-ab8664745 #1 pc : ufshcd_release_scsi_cmd+0x30/0x46c lr : __ufshcd_transfer_req_compl+0x4fc/0x9c0 Call trace: ufshcd_release_scsi_cmd+0x30/0x46c __ufshcd_transfer_req_compl+0x4fc/0x9c0 ufshcd_poll+0xf0/0x208 ufshcd_sl_intr+0xb8/0xf0 ufshcd_intr+0x168/0x2f4 __handle_irq_event_percpu+0xa0/0x30c handle_irq_event+0x84/0x178 handle_fasteoi_irq+0x150/0x2e8 __handle_domain_irq+0x114/0x1e4 gic_handle_irq.31846+0x58/0x300 el1_irq+0xe4/0x1c0 cpuidle_enter_state+0x3ac/0x8c4 do_idle+0x2fc/0x55c cpu_startup_entry+0x84/0x90 kernel_init+0x0/0x310 start_kernel+0x0/0x608 start_kernel+0x4ec/0x608 Signed-off-by: Bart Van Assche <bvanassche@acm.org> --- drivers/scsi/ufs/ufshcd.c | 20 +++++++++++++------- 1 file changed, 13 insertions(+), 7 deletions(-)