Retry READ CAPACITY(10)/(16) with good status but no data
cover Retry READ CAPACITY(10)/(16) with good status but no data
We encountered a SCSI device that responded to the initial READ CAPACITY command with a good status, but no data was transferred. This caused a sudden change of the device capacity to zero when the device was rescanned, for no obvious reason.
This patch series changes read_capacity_10() and read_capacity_16() in sd.c to retry the command up to 3 times in an attempt to get valid capacity information. A message is logged if this is ultimately unsuccessful.
There are some predecessor patches, one from a patch in a series by Mike Christie which changes read_capacity_16() to use the scsi_failures mechanism (which did not eventually get merged), this makes the changes here much more similar for both the read_capacity_10 and read_capacity_16() case. Another patch corrects a potential use of an uninitialized variable, and a third one removes a check for -EOVERFLOW that hasn't been needed since commit 72deb455b5ec ("block: remove CONFIG_LBDAF").
The final patch to scsi_debug is allow insertion of the fault to test this change.
Ewan D. Milne (4):
scsi: sd: Avoid passing potentially uninitialized "sense_valid" to
read_capacity_error()
scsi: sd: Remove checks for -EOVERFLOW in sd_read_capacity()
scsi: sd: Check for and retry in case of READ_CAPCITY(10)/(16)
returning no data
scsi: scsi_debug: Add option to suppress returned data but return good
status
Mike Christie (1):
scsi: sd: Have scsi-ml retry read_capacity_16 errors
drivers/scsi/scsi_debug.c | 38 ++++++--- drivers/scsi/sd.c | 173 +++++++++++++++++++++++++++----------- 2 files changed, 151 insertions(+), 60 deletions(-)
Signature · 1 line
2.47.1
PATCH 1/5 scsi: sd: Have scsi-ml retry read_capacity_16 errors
From: Mike Christie <[email protected]>
This has read_capacity_16 have scsi-ml retry errors instead of driving them itself.
There are 2 behavior changes with this patch:
1. There is one behavior change where we no longer retry when scsi_execute_cmd returns < 0, but we should be ok. We don't need to retry for failures like the queue being removed, and for the case where there are no tags/reqs since the block layer waits/retries for us. For possible memory allocation failures from blk_rq_map_kern we use GFP_NOIO, so retrying will probably not help.
2. For the specific UAs we checked for and retried, we would get READ_CAPACITY_RETRIES_ON_RESET retries plus whatever retries were left from the main loop's retries. Each UA now gets READ_CAPACITY_RETRIES_ON_RESET reties, and the other errors get up to 3 retries. This is most likely ok, because READ_CAPACITY_RETRIES_ON_RESET is already 10 and is not based on anything specific like a spec or device, so the extra 3 we got from the main loop was probably just an accident and is not going to help.
--- drivers/scsi/sd.c | 108 ++++++++++++++++++++++++++++++---------------- 1 file changed, 71 insertions(+), 37 deletions(-)
drivers/scsi/sd.c +71 -37
@@ -2626,55 +2626,89 @@ static void read_capacity_error(struct scsi_disk *sdkp, struct scsi_device *sdp,static int read_capacity_16 (struct scsi_disk *sdkp,struct scsi_device *sdp,struct queue_limits *lim,unsigned char *buffer) { -unsigned char cmd[16 ]; -struct scsi_sense_hdr sshdr; -const struct scsi_exec_args exec_args = { - .sshdr = &sshdr, +static const u8 cmd[16 ] = { + [0 ] =SERVICE_ACTION_IN_16 , + [1 ] =SAI_READ_CAPACITY_16 , + [13 ] =RC16_LEN , }; +struct scsi_sense_hdr sshdr;int sense_valid =0 ;int the_result; -int retries =3 , reset_retries =READ_CAPACITY_RETRIES_ON_RESET ;unsigned int alignment;unsigned long long lba;unsigned sector_size; +struct scsi_failure failure_defs[] = { +/* +* Do not retry Invalid Command Operation Code or Invalid +* Field in CDB. +*/ + { + .sense =ILLEGAL_REQUEST , + .asc =0x20 , + .result =SAM_STAT_CHECK_CONDITION , + }, + { + .sense =ILLEGAL_REQUEST , + .asc =0x24 , + .result =SAM_STAT_CHECK_CONDITION , + }, +/* Do not retry Medium Not Present */ + { + .sense =UNIT_ATTENTION , + .asc =0x3A , + .result =SAM_STAT_CHECK_CONDITION , + }, + { + .sense =NOT_READY , + .asc =0x3A , + .result =SAM_STAT_CHECK_CONDITION , + }, +/* Device reset might occur several times so retry a lot */ + { + .sense =UNIT_ATTENTION , + .asc =0x29 , + .allowed =READ_CAPACITY_RETRIES_ON_RESET , + .result =SAM_STAT_CHECK_CONDITION , + }, +/* Any other error not listed above retry 3 times */ + { + .result =SCMD_FAILURE_RESULT_ANY , + .allowed =3 , + }, + {} + }; +struct scsi_failures failures = { + .failure_definitions = failure_defs, + }; +const struct scsi_exec_args exec_args = { + .sshdr = &sshdr, + .failures = &failures, + };if (sdp->no_read_capacity_16)return -EINVAL ; -do { -memset (cmd,0 ,16 ); - cmd[0 ] =SERVICE_ACTION_IN_16 ; - cmd[1 ] =SAI_READ_CAPACITY_16 ; - cmd[13 ] =RC16_LEN ; -memset (buffer,0 ,RC16_LEN ); - - the_result =scsi_execute_cmd (sdp, cmd,REQ_OP_DRV_IN , - buffer,RC16_LEN ,SD_TIMEOUT , - sdkp->max_retries, &exec_args); -if (the_result >0 ) { -if (media_not_present (sdkp, &sshdr)) -return -ENODEV ; +memset (buffer,0 ,RC16_LEN ); - sense_valid =scsi_sense_valid (&sshdr); -if (sense_valid && - sshdr.sense_key ==ILLEGAL_REQUEST && - (sshdr.asc ==0x20 || sshdr.asc ==0x24 ) && - sshdr.ascq ==0x00 ) -/* Invalid Command Operation Code or -* Invalid Field in CDB, just retry -* silently with RC10 */ -return -EINVAL ; -if (sense_valid && - sshdr.sense_key ==UNIT_ATTENTION && - sshdr.asc ==0x29 && sshdr.ascq ==0x00 ) -/* Device reset might occur several times, -* give it one more chance */ -if (--reset_retries >0 ) -continue ; - } - retries--; + the_result =scsi_execute_cmd (sdp, cmd,REQ_OP_DRV_IN , buffer, +RC16_LEN ,SD_TIMEOUT , sdkp->max_retries, + &exec_args); - }while (the_result && retries); +if (the_result >0 ) { +if (media_not_present (sdkp, &sshdr)) +return -ENODEV ; + + sense_valid =scsi_sense_valid (&sshdr); +if (sense_valid && sshdr.sense_key ==ILLEGAL_REQUEST && + (sshdr.asc ==0x20 || sshdr.asc ==0x24 ) && + sshdr.ascq ==0x00 ) { +/* +* Invalid Command Operation Code or Invalid Field in +* CDB, just retry silently with RC10 +*/ +return -EINVAL ; + } + }if (the_result) {sd_print_result (sdkp,"Read Capacity(16) failed" , the_result);
Signature · 1 line
2.47.1
Re: [PATCH 1/5] scsi: sd: Have scsi-ml retry read_capacity_16 errors
Quoted 18 lines · Ewan D. Milne
This has
read_capacity_16have scsi-ml retry errors instead of driving them itself.There are 2 behavior changes with this patch:
1. There is one behavior change where we no longer retry whenscsi_execute_cmdreturns < 0, but we should be ok. We don't need to retry for failures like the queue being removed, and for the case where there are no tags/reqs since the block layer waits/retries for us. For possible memory allocation failures fromblk_rq_map_kernwe useGFP_NOIO, so retrying will probably not help.
2. For the specific UAs we checked for and retried, we would getREAD_CAPACITY_RETRIES_ON_RESETretries plus whatever retries were left from the main loop's retries. Each UA now getsREAD_CAPACITY_RETRIES_ON_RESETreties, and the other errors get up to 3 retries. This is most likely ok, becauseREAD_CAPACITY_RETRIES_ON_RESETis already 10 and is not based on anything specific like a spec or device, so the extra 3 we got from the main loop was probably just an accident and is not going to help.
Is this the same patch as https://lore.kernel.org/linux-scsi/[email protected]/? If so, does John Garry's Reviewed-by still apply?
Thanks,
Bart.
Re: [PATCH 1/5] scsi: sd: Have scsi-ml retry read_capacity_16 errors
In reply to Ewan D. MilneQuoted 22 lines · Ewan D. Milne
drivers/scsi/sd.c +22
-0+/* +* Do not retry Invalid Command Operation Code or Invalid +* Field in CDB. +*/ + { + .sense = ILLEGAL_REQUEST, + .asc = 0x20, + .result = SAM_STAT_CHECK_CONDITION, + }, + { + .sense = ILLEGAL_REQUEST, + .asc = 0x24, + .result = SAM_STAT_CHECK_CONDITION, + }, + /* Do not retry Medium Not Present */ + { + .sense = UNIT_ATTENTION, + .asc = 0x3A, + .result = SAM_STAT_CHECK_CONDITION, + }, + { + .sense = NOT_READY,
drivers/scsi/sd.c static void read_capacity_error(struct scsi_disk *sdkp, struct scsi_device *sdp, · line 2672
Quoted 10 lines
drivers/scsi/sd.c +10
-0+ .asc = 0x3A, + .result = SAM_STAT_CHECK_CONDITION, + }, +/* Device reset might occur several times so retry a lot */ + { + .sense = UNIT_ATTENTION, + .asc = 0x29, + .allowed = READ_CAPACITY_RETRIES_ON_RESET, + .result = SAM_STAT_CHECK_CONDITION, + },
For the first, second and fifth array elements above: leaving out .ascq is the same as requiring that the ASCQ value is zero. I prefer to make this explicit by adding ".ascq = 0," in these array elements.
Additionally, media_not_present() doesn't check the ASCQ value while the above array only accepts ASCQ == 0 if ASC == 0x3a. Please either mention this behavior change in the patch description or add the following in the third and fourth array elements:
".ascq = SCMD_FAILURE_ASCQ_ANY,".
drivers/scsi/sd.c static void read_capacity_error(struct scsi_disk *sdkp, struct scsi_device *sdp, · line 2691
Quoted 1 line
drivers/scsi/sd.c +1
-0+memset (buffer,0 ,RC16_LEN );
Isn't the preferred style "memset(buffer, 0, ARRAY_SIZE(buffer))"? This makes it easier for readers to verify that the third argument is correct.
Otherwise this patch looks good to me.
Thanks,
Bart.
PATCH 2/5 scsi: sd: Avoid passing potentially uninitialized "sense_valid" to read_capacity_error()
In reply to Ewan D. Milneread_capacity_10() sets "sense_valid" in a different conditional statement prior to calling read_capacity_error(), and does not use this value otherwise. Move the call to scsi_sense_valid() to read_capacity_error() instead of passing it as a parameter from read_capacity_16() and read_capacity_10().
--- drivers/scsi/sd.c | 12 +++++------- 1 file changed, 5 insertions(+), 7 deletions(-)
drivers/scsi/sd.c +5 -7
@@ -2592,9 +2592,10 @@ static void sd_config_protection(struct scsi_disk *sdkp, }static void read_capacity_error (struct scsi_disk *sdkp,struct scsi_device *sdp, -struct scsi_sense_hdr *sshdr,int sense_valid, -int the_result) +struct scsi_sense_hdr *sshdr,int the_result) { +bool sense_valid =scsi_sense_valid (sshdr); +if (sense_valid)sd_print_sense_hdr (sdkp, sshdr);else
@@ -2712,7 +2713,7 @@ static int read_capacity_16(struct scsi_disk *sdkp, struct scsi_device *sdp,if (the_result) {sd_print_result (sdkp,"Read Capacity(16) failed" , the_result); -read_capacity_error (sdkp, sdp, &sshdr, sense_valid, the_result); +read_capacity_error (sdkp, sdp, &sshdr, the_result);return -EINVAL ; }
@@ -2786,7 +2787,6 @@ static int read_capacity_10(struct scsi_disk *sdkp, struct scsi_device *sdp, .sshdr = &sshdr, .failures = &failures, }; -int sense_valid =0 ;int the_result;sector_t lba;unsigned sector_size;
@@ -2798,15 +2798,13 @@ static int read_capacity_10(struct scsi_disk *sdkp, struct scsi_device *sdp, &exec_args);if (the_result >0 ) { - sense_valid =scsi_sense_valid (&sshdr); -if (media_not_present (sdkp, &sshdr))return -ENODEV ; }if (the_result) {sd_print_result (sdkp,"Read Capacity(10) failed" , the_result); -read_capacity_error (sdkp, sdp, &sshdr, sense_valid, the_result); +read_capacity_error (sdkp, sdp, &sshdr, the_result);return -EINVAL ; }
Signature · 1 line
2.47.1
Re: [PATCH 2/5] scsi: sd: Avoid passing potentially uninitialized "sense_valid" to read_capacity_error()
drivers/scsi/sd.c static int read_capacity_16(struct scsi_disk *sdkp, struct scsi_device *sdp, · line 2719
Quoted 9 lines · Ewan D. Milne
drivers/scsi/sd.c +1
-1@@ -2712,7 +2713,7 @@ static int read_capacity_16(struct scsi_disk *sdkp, struct scsi_device *sdp,if (the_result) {sd_print_result (sdkp,"Read Capacity(16) failed" , the_result); -read_capacity_error (sdkp, sdp, &sshdr, sense_valid, the_result); +read_capacity_error (sdkp, sdp, &sshdr, the_result);return -EINVAL ; }
Readability of read_capacity_16() would improve if the 'sense_valid' declaration would be moved into the only scope where it is used. Otherwise this patch looks good to me.
Bart.
PATCH 3/5 scsi: sd: Remove checks for -EOVERFLOW in sd_read_capacity()
In reply to Ewan D. MilneRemove checks for -EOVERFLOW in sd_read_capacity() because this value has not been returned to it since commit 72deb455b5ec ("block: remove CONFIG_LBDAF").
--- drivers/scsi/sd.c | 4 ---- 1 file changed, 4 deletions(-)
drivers/scsi/sd.c +0 -4
@@ -2850,8 +2850,6 @@ sd_read_capacity(struct scsi_disk *sdkp, struct queue_limits *lim,if (sd_try_rc16_first (sdp)) { sector_size =read_capacity_16 (sdkp, sdp, lim, buffer); -if (sector_size == -EOVERFLOW ) - goto got_data;if (sector_size == -ENODEV )return ;if (sector_size <0 )
@@ -2860,8 +2858,6 @@ sd_read_capacity(struct scsi_disk *sdkp, struct queue_limits *lim,return ; }else { sector_size =read_capacity_10 (sdkp, sdp, buffer); -if (sector_size == -EOVERFLOW ) - goto got_data;if (sector_size <0 )return ; if ((sizeof(sdkp->capacity) > 4) &&
Signature · 1 line
2.47.1
Re: [PATCH 3/5] scsi: sd: Remove checks for -EOVERFLOW in sd_read_capacity()
Quoted 2 lines · Ewan D. Milne
Remove checks for
-EOVERFLOWinsd_read_capacity()because this value has not been returned to it since commit72deb455b5ec("block: removeCONFIG_LBDAF").
PATCH 4/5 scsi: sd: Check for and retry in case of READ_CAPCITY(10)/(16) returning no data
In reply to Ewan D. Milnesd_read_capacity_10() and sd_read_capacity_16() do not check for underflow and can extract invalid (e.g. zero) data when a malfunctioning device does not actually transfer any data, but returnes a good status otherwise. Check for this and retry, and log a message and return -EINVAL if we can't get the capacity information.
We encountered a device that did this once but returned good data afterwards.
See similar commit 5cd3bbfad088 ("[SCSI] retry with missing data for INQUIRY")
--- drivers/scsi/sd.c | 69 +++++++++++++++++++++++++++++++++++++++-------- 1 file changed, 58 insertions(+), 11 deletions(-)
drivers/scsi/sd.c +58 -11
@@ -2633,6 +2633,7 @@ static int read_capacity_16(struct scsi_disk *sdkp, struct scsi_device *sdp, [13] = RC16_LEN, };struct scsi_sense_hdr sshdr; +int count, resid;int sense_valid =0 ;int the_result;unsigned int alignment;
@@ -2683,17 +2684,31 @@ static int read_capacity_16(struct scsi_disk *sdkp, struct scsi_device *sdp, };const struct scsi_exec_args exec_args = { .sshdr = &sshdr, + .resid = &resid, .failures = &failures, };if (sdp->no_read_capacity_16)return -EINVAL ; -memset (buffer,0 ,RC16_LEN ); +for (count =0 ; count <3 ; ++count) { +memset (buffer,0 ,RC16_LEN ); - the_result =scsi_execute_cmd (sdp, cmd,REQ_OP_DRV_IN , buffer, -RC16_LEN ,SD_TIMEOUT , sdkp->max_retries, - &exec_args); + the_result =scsi_execute_cmd (sdp, cmd,REQ_OP_DRV_IN , + buffer,RC16_LEN ,SD_TIMEOUT , + sdkp->max_retries, &exec_args); + +if (the_result ==0 ) { +/* +* if nothing was transferred, we try +* again. It's a workaround for a broken +* device. +*/ +if (resid ==RC16_LEN ) +continue ; + } +break ; + }if (the_result >0 ) {if (media_not_present (sdkp, &sshdr))
@@ -2717,6 +2732,12 @@ static int read_capacity_16(struct scsi_disk *sdkp, struct scsi_device *sdp,return -EINVAL ; } +if (resid ==RC16_LEN ) { +sd_printk (KERN_ERR , sdkp, +"Read Capacity(16) returned good status but no data" ); +return -EINVAL ; + } + sector_size =get_unaligned_be32 (&buffer[8 ]); lba =get_unaligned_be64 (&buffer[0 ]);
@@ -2749,11 +2770,20 @@ static int read_capacity_16(struct scsi_disk *sdkp, struct scsi_device *sdp,return sector_size; } +#define RC10_LEN 8 +#if RC10_LEN >SD_BUF_SIZE +#error RC10_LEN must not be more than SD_BUF_SIZE +#endif +static int read_capacity_10 (struct scsi_disk *sdkp,struct scsi_device *sdp,unsigned char *buffer) {static const u8 cmd[10 ] = {READ_CAPACITY };struct scsi_sense_hdr sshdr; +int count, resid; +int the_result; +sector_t lba; +unsigned sector_size;struct scsi_failure failure_defs[] = {/* Do not retry Medium Not Present */ {
@@ -2785,17 +2815,28 @@ static int read_capacity_10(struct scsi_disk *sdkp, struct scsi_device *sdp, };const struct scsi_exec_args exec_args = { .sshdr = &sshdr, + .resid = &resid, .failures = &failures, }; -int the_result; -sector_t lba; -unsigned sector_size; -memset (buffer,0 ,8 ); +for (count =0 ; count <3 ; ++count) { +memset (buffer,0 ,RC10_LEN ); - the_result =scsi_execute_cmd (sdp, cmd,REQ_OP_DRV_IN , buffer, -8 ,SD_TIMEOUT , sdkp->max_retries, - &exec_args); + the_result =scsi_execute_cmd (sdp, cmd,REQ_OP_DRV_IN , + buffer,RC10_LEN ,SD_TIMEOUT , + sdkp->max_retries, &exec_args); + +if (the_result ==0 ) { +/* +* if nothing was transferred, we try +* again. It's a workaround for a broken +* device. +*/ +if (resid ==RC10_LEN ) +continue ; + } +break ; + }if (the_result >0 ) {if (media_not_present (sdkp, &sshdr))
@@ -2808,6 +2849,12 @@ static int read_capacity_10(struct scsi_disk *sdkp, struct scsi_device *sdp,return -EINVAL ; } +if (resid ==RC10_LEN ) { +sd_printk (KERN_ERR , sdkp, +"Read Capacity(10) returned good status but no data" ); +return -EINVAL ; + } + sector_size =get_unaligned_be32 (&buffer[4 ]); lba =get_unaligned_be32 (&buffer[0 ]);
Signature · 1 line
2.47.1
Re: [PATCH 4/5] scsi: sd: Check for and retry in case of READ_CAPCITY(10)/(16) returning no data
Quoted 3 lines · Ewan D. Milne
sd_read_capacity_10()andsd_read_capacity_16()do not check for underflow and can extract invalid (e.g. zero) data when a malfunctioning device does not actually transfer any data, but returnes a good status otherwise.
returns?
drivers/scsi/sd.c static int read_capacity_16(struct scsi_disk *sdkp, struct scsi_device *sdp, · line 2709
Quoted 9 lines
drivers/scsi/sd.c +9
-0+if (the_result ==0 ) { +/* +* if nothing was transferred, we try +* again. It's a workaround for a broken +* device. +*/ +if (resid ==RC16_LEN ) +continue ; + }
Please combine the above two if-statements into a single if-statement.
drivers/scsi/sd.c static int read_capacity_16(struct scsi_disk *sdkp, struct scsi_device *sdp, · line 2776
Quoted 4 lines
drivers/scsi/sd.c +4
-0+#define RC10_LEN 8 +#if RC10_LEN >SD_BUF_SIZE +#error RC10_LEN must not be more than SD_BUF_SIZE +#endif
Please pass the 'buffer' length as an argument to read_capacity_10(). That will make it easier to spot an inconsistent buffer size value by auditing the caller of this function (I am aware that the above code follows the style of the read_capacity_16() function).
drivers/scsi/sd.c static int read_capacity_10(struct scsi_disk *sdkp, struct scsi_device *sdp, · line 2837
Quoted 9 lines
drivers/scsi/sd.c +9
-0+if (the_result ==0 ) { +/* +* if nothing was transferred, we try +* again. It's a workaround for a broken +* device. +*/ +if (resid ==RC10_LEN ) +continue ; + }
Also here, please combine the two if-statements into a single if-statement.
Thanks,
Bart.
PATCH 5/5 scsi: scsi_debug: Add option to suppress returned data but return good status
In reply to Ewan D. MilneThis is used to test the earlier read_capacity_10()/16() retry patch.
--- drivers/scsi/scsi_debug.c | 38 +++++++++++++++++++++++++++----------- 1 file changed, 27 insertions(+), 11 deletions(-)
drivers/scsi/scsi_debug.c +27 -11
@@ -230,6 +230,7 @@ struct tape_block {#define SDEBUG_OPT_NO_CDB_NOISE 0x4000#define SDEBUG_OPT_HOST_BUSY 0x8000#define SDEBUG_OPT_CMD_ABORT 0x10000 +#define SDEBUG_OPT_NO_DATA 0x20000#define SDEBUG_OPT_ALL_NOISE (SDEBUG_OPT_NOISE | SDEBUG_OPT_Q_NOISE | \ SDEBUG_OPT_RESET_NOISE)#define SDEBUG_OPT_ALL_INJECTING (SDEBUG_OPT_RECOVERED_ERR | \
@@ -1641,10 +1642,17 @@ static int fill_from_dev_buffer(struct scsi_cmnd *scp, unsigned char *arr,if (scp->sc_data_direction !=DMA_FROM_DEVICE )return DID_ERROR <<16 ; - act_len =sg_copy_from_buffer (sdb->table.sgl, sdb->table.nents, - arr, arr_len); -scsi_set_resid (scp,scsi_bufflen (scp) - act_len); - +/* +* Conditionally suppress DATA IN transfer and leave resid set to bufflen. +*/ +if (unlikely ((sdebug_opts &SDEBUG_OPT_NO_DATA ) && +atomic_read (&sdeb_inject_pending))) { +scsi_set_resid (scp,scsi_bufflen (scp)); + }else { + act_len =sg_copy_from_buffer (sdb->table.sgl, sdb->table.nents, + arr, arr_len); +scsi_set_resid (scp,scsi_bufflen (scp) - act_len); + }return 0 ; }
@@ -1665,13 +1673,21 @@ static int p_fill_from_dev_buffer(struct scsi_cmnd *scp, const void *arr,if (scp->sc_data_direction !=DMA_FROM_DEVICE )return DID_ERROR <<16 ; - act_len =sg_pcopy_from_buffer (sdb->table.sgl, sdb->table.nents, - arr, arr_len, skip); -pr_debug ("%s: off_dst=%u, scsi_bufflen=%u, act_len=%u, resid=%d\n" , - __func__, off_dst,scsi_bufflen (scp), act_len, -scsi_get_resid (scp)); - n =scsi_bufflen (scp) - (off_dst + act_len); -scsi_set_resid (scp,min_t (u32,scsi_get_resid (scp), n)); +/* +* Conditionally suppress DATA IN transfer and leave resid set to bufflen. +*/ +if (unlikely ((sdebug_opts &SDEBUG_OPT_NO_DATA ) && +atomic_read (&sdeb_inject_pending))) { +scsi_set_resid (scp,scsi_bufflen (scp)); + }else { + act_len =sg_pcopy_from_buffer (sdb->table.sgl, sdb->table.nents, + arr, arr_len, skip); +pr_debug ("%s: off_dst=%u, scsi_bufflen=%u, act_len=%u, resid=%d\n" , + __func__, off_dst,scsi_bufflen (scp), act_len, +scsi_get_resid (scp)); + n =scsi_bufflen (scp) - (off_dst + act_len); +scsi_set_resid (scp,min_t (u32,scsi_get_resid (scp), n)); + }return 0 ; }
Signature · 1 line
2.47.1
Re: [PATCH 5/5] scsi: scsi_debug: Add option to suppress returned data but return good status
Quoted 52 lines · Ewan D. Milne
drivers/scsi/scsi_debug.c +19
-11@@ -230,6 +230,7 @@ struct tape_block {#define SDEBUG_OPT_NO_CDB_NOISE 0x4000#define SDEBUG_OPT_HOST_BUSY 0x8000#define SDEBUG_OPT_CMD_ABORT 0x10000 +#define SDEBUG_OPT_NO_DATA 0x20000#define SDEBUG_OPT_ALL_NOISE (SDEBUG_OPT_NOISE | SDEBUG_OPT_Q_NOISE | \ SDEBUG_OPT_RESET_NOISE)#define SDEBUG_OPT_ALL_INJECTING (SDEBUG_OPT_RECOVERED_ERR | \@@ -1641,10 +1642,17 @@ static int fill_from_dev_buffer(struct scsi_cmnd *scp, unsigned char *arr,if (scp->sc_data_direction !=DMA_FROM_DEVICE )return DID_ERROR <<16 ; - act_len =sg_copy_from_buffer (sdb->table.sgl, sdb->table.nents, - arr, arr_len); -scsi_set_resid (scp,scsi_bufflen (scp) - act_len); - +/* +* Conditionally suppress DATA IN transfer and leave resid set to bufflen. +*/ +if (unlikely ((sdebug_opts &SDEBUG_OPT_NO_DATA ) && +atomic_read (&sdeb_inject_pending))) { +scsi_set_resid (scp,scsi_bufflen (scp)); + }else { + act_len =sg_copy_from_buffer (sdb->table.sgl, sdb->table.nents, + arr, arr_len); +scsi_set_resid (scp,scsi_bufflen (scp) - act_len); + }return 0 ; }@@ -1665,13 +1673,21 @@ static int p_fill_from_dev_buffer(struct scsi_cmnd *scp, const void *arr,if (scp->sc_data_direction !=DMA_FROM_DEVICE )return DID_ERROR <<16 ; - act_len =sg_pcopy_from_buffer (sdb->table.sgl, sdb->table.nents, - arr, arr_len, skip); -pr_debug ("%s: off_dst=%u, scsi_bufflen=%u, act_len=%u, resid=%d\n" , - __func__, off_dst,scsi_bufflen (scp), act_len, -scsi_get_resid (scp)); - n =scsi_bufflen (scp) - (off_dst + act_len); -scsi_set_resid (scp,min_t (u32,scsi_get_resid (scp), n)); +/* +* Conditionally suppress DATA IN transfer and leave resid set to bufflen. +*/ +if (unlikely ((sdebug_opts &SDEBUG_OPT_NO_DATA ) && +atomic_read (&sdeb_inject_pending))) { +scsi_set_resid (scp,scsi_bufflen (scp)); + }else {
drivers/scsi/scsi_debug.c static int p_fill_from_dev_buffer(struct scsi_cmnd *scp, const void *arr, · line 1692
Quoted 10 lines
drivers/scsi/scsi_debug.c +8
-0+ act_len =sg_pcopy_from_buffer (sdb->table.sgl, sdb->table.nents, + arr, arr_len, skip); +pr_debug ("%s: off_dst=%u, scsi_bufflen=%u, act_len=%u, resid=%d\n" , + __func__, off_dst,scsi_bufflen (scp), act_len, +scsi_get_resid (scp)); + n =scsi_bufflen (scp) - (off_dst + act_len); +scsi_set_resid (scp,min_t (u32,scsi_get_resid (scp), n)); + }return 0 ; }
Here and above, please combine the two scsi_set_resid() calls into a single scsi_set_resid() call, e.g. by setting act_len to zero when the DATA IN transfer is suppressed.
Thanks,
Bart.
Gentle ping for reviews.
Thanks.
-Ewan
Quoted 39 lines · Ewan D. Milne
We encountered a SCSI device that responded to the initial READ CAPACITY command with a good status, but no data was transferred. This caused a sudden change of the device capacity to zero when the device was rescanned, for no obvious reason.
This patch series changes
read_capacity_10()andread_capacity_16()insd.cto retry the command up to 3 times in an attempt to get valid capacity information. A message is logged if this is ultimately unsuccessful.There are some predecessor patches, one from a patch in a series by Mike Christie which changes
read_capacity_16()to use thescsi_failuresmechanism (which did not eventually get merged), this makes the changes here much more similar for both theread_capacity_10andread_capacity_16()case. Another patch corrects a potential use of an uninitialized variable, and a third one removes a check for-EOVERFLOWthat hasn't been needed since commit72deb455b5ec("block: removeCONFIG_LBDAF").The final patch to
scsi_debugis allow insertion of the fault to test this change.Ewan D. Milne (4):
scsi: sd: Avoid passing potentially uninitialized "sense_valid" to
read_capacity_error()
scsi: sd: Remove checks for-EOVERFLOWinsd_read_capacity()
scsi: sd: Check for and retry in case of READ_CAPCITY(10)/(16)
returning no data
scsi:scsi_debug: Add option to suppress returned data but return good
statusMike Christie (1):
scsi: sd: Have scsi-ml retryread_capacity_16errorsdrivers/scsi/scsi_debug.c | 38 ++++++--- drivers/scsi/sd.c | 173 +++++++++++++++++++++++++++----------- 2 files changed, 151 insertions(+), 60 deletions(-)--
2.47.1
Ewan,
Quoted 1 line
Gentle ping for reviews.
I will take another look once the merge window has closed. But no objections to the general approach.
Since this constitutes a core change, I would like another reviewer or two...
Signature · 1 line
Martin K. Petersen