From 1b9a9ab78b0ab79dc1f0ddd5fbed7833ec7de3a4 Mon Sep 17 00:00:00 2001 From: Mike Christie Date: Sun, 5 Jun 2016 14:32:18 -0500 Subject: blktrace: use op accessors Have blktrace use the req/bio op accessor to get the REQ_OP. Signed-off-by: Mike Christie Reviewed-by: Christoph Hellwig Reviewed-by: Hannes Reinecke Signed-off-by: Jens Axboe --- kernel/trace/blktrace.c | 62 ++++++++++++++++++++++++++++--------------------- 1 file changed, 36 insertions(+), 26 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/blktrace.c b/kernel/trace/blktrace.c index 9aef8654e90d..2d16fad519b2 100644 --- a/kernel/trace/blktrace.c +++ b/kernel/trace/blktrace.c @@ -199,7 +199,8 @@ static const u32 ddir_act[2] = { BLK_TC_ACT(BLK_TC_READ), * blk_io_trace structure and places it in a per-cpu subbuffer. */ static void __blk_add_trace(struct blk_trace *bt, sector_t sector, int bytes, - int rw, u32 what, int error, int pdu_len, void *pdu_data) + int op, int op_flags, u32 what, int error, int pdu_len, + void *pdu_data) { struct task_struct *tsk = current; struct ring_buffer_event *event = NULL; @@ -214,13 +215,14 @@ static void __blk_add_trace(struct blk_trace *bt, sector_t sector, int bytes, if (unlikely(bt->trace_state != Blktrace_running && !blk_tracer)) return; - what |= ddir_act[rw & WRITE]; - what |= MASK_TC_BIT(rw, SYNC); - what |= MASK_TC_BIT(rw, RAHEAD); - what |= MASK_TC_BIT(rw, META); - what |= MASK_TC_BIT(rw, DISCARD); - what |= MASK_TC_BIT(rw, FLUSH); - what |= MASK_TC_BIT(rw, FUA); + what |= ddir_act[op_is_write(op) ? WRITE : READ]; + what |= MASK_TC_BIT(op_flags, SYNC); + what |= MASK_TC_BIT(op_flags, RAHEAD); + what |= MASK_TC_BIT(op_flags, META); + what |= MASK_TC_BIT(op_flags, FLUSH); + what |= MASK_TC_BIT(op_flags, FUA); + if (op == REQ_OP_DISCARD) + what |= BLK_TC_ACT(BLK_TC_DISCARD); pid = tsk->pid; if (act_log_check(bt, what, sector, pid)) @@ -708,11 +710,11 @@ static void blk_add_trace_rq(struct request_queue *q, struct request *rq, if (rq->cmd_type == REQ_TYPE_BLOCK_PC) { what |= BLK_TC_ACT(BLK_TC_PC); - __blk_add_trace(bt, 0, nr_bytes, rq->cmd_flags, + __blk_add_trace(bt, 0, nr_bytes, req_op(rq), rq->cmd_flags, what, rq->errors, rq->cmd_len, rq->cmd); } else { what |= BLK_TC_ACT(BLK_TC_FS); - __blk_add_trace(bt, blk_rq_pos(rq), nr_bytes, + __blk_add_trace(bt, blk_rq_pos(rq), nr_bytes, req_op(rq), rq->cmd_flags, what, rq->errors, 0, NULL); } } @@ -770,7 +772,7 @@ static void blk_add_trace_bio(struct request_queue *q, struct bio *bio, return; __blk_add_trace(bt, bio->bi_iter.bi_sector, bio->bi_iter.bi_size, - bio->bi_rw, what, error, 0, NULL); + bio_op(bio), bio->bi_rw, what, error, 0, NULL); } static void blk_add_trace_bio_bounce(void *ignore, @@ -818,7 +820,8 @@ static void blk_add_trace_getrq(void *ignore, struct blk_trace *bt = q->blk_trace; if (bt) - __blk_add_trace(bt, 0, 0, rw, BLK_TA_GETRQ, 0, 0, NULL); + __blk_add_trace(bt, 0, 0, rw, 0, BLK_TA_GETRQ, 0, 0, + NULL); } } @@ -833,7 +836,7 @@ static void blk_add_trace_sleeprq(void *ignore, struct blk_trace *bt = q->blk_trace; if (bt) - __blk_add_trace(bt, 0, 0, rw, BLK_TA_SLEEPRQ, + __blk_add_trace(bt, 0, 0, rw, 0, BLK_TA_SLEEPRQ, 0, 0, NULL); } } @@ -843,7 +846,7 @@ static void blk_add_trace_plug(void *ignore, struct request_queue *q) struct blk_trace *bt = q->blk_trace; if (bt) - __blk_add_trace(bt, 0, 0, 0, BLK_TA_PLUG, 0, 0, NULL); + __blk_add_trace(bt, 0, 0, 0, 0, BLK_TA_PLUG, 0, 0, NULL); } static void blk_add_trace_unplug(void *ignore, struct request_queue *q, @@ -860,7 +863,7 @@ static void blk_add_trace_unplug(void *ignore, struct request_queue *q, else what = BLK_TA_UNPLUG_TIMER; - __blk_add_trace(bt, 0, 0, 0, what, 0, sizeof(rpdu), &rpdu); + __blk_add_trace(bt, 0, 0, 0, 0, what, 0, sizeof(rpdu), &rpdu); } } @@ -874,8 +877,9 @@ static void blk_add_trace_split(void *ignore, __be64 rpdu = cpu_to_be64(pdu); __blk_add_trace(bt, bio->bi_iter.bi_sector, - bio->bi_iter.bi_size, bio->bi_rw, BLK_TA_SPLIT, - bio->bi_error, sizeof(rpdu), &rpdu); + bio->bi_iter.bi_size, bio_op(bio), bio->bi_rw, + BLK_TA_SPLIT, bio->bi_error, sizeof(rpdu), + &rpdu); } } @@ -907,7 +911,7 @@ static void blk_add_trace_bio_remap(void *ignore, r.sector_from = cpu_to_be64(from); __blk_add_trace(bt, bio->bi_iter.bi_sector, bio->bi_iter.bi_size, - bio->bi_rw, BLK_TA_REMAP, bio->bi_error, + bio_op(bio), bio->bi_rw, BLK_TA_REMAP, bio->bi_error, sizeof(r), &r); } @@ -940,7 +944,7 @@ static void blk_add_trace_rq_remap(void *ignore, r.sector_from = cpu_to_be64(from); __blk_add_trace(bt, blk_rq_pos(rq), blk_rq_bytes(rq), - rq_data_dir(rq), BLK_TA_REMAP, !!rq->errors, + rq_data_dir(rq), 0, BLK_TA_REMAP, !!rq->errors, sizeof(r), &r); } @@ -965,10 +969,10 @@ void blk_add_driver_data(struct request_queue *q, return; if (rq->cmd_type == REQ_TYPE_BLOCK_PC) - __blk_add_trace(bt, 0, blk_rq_bytes(rq), 0, + __blk_add_trace(bt, 0, blk_rq_bytes(rq), 0, 0, BLK_TA_DRV_DATA, rq->errors, len, data); else - __blk_add_trace(bt, blk_rq_pos(rq), blk_rq_bytes(rq), 0, + __blk_add_trace(bt, blk_rq_pos(rq), blk_rq_bytes(rq), 0, 0, BLK_TA_DRV_DATA, rq->errors, len, data); } EXPORT_SYMBOL_GPL(blk_add_driver_data); @@ -1769,21 +1773,27 @@ void blk_dump_cmd(char *buf, struct request *rq) } } -void blk_fill_rwbs(char *rwbs, u32 rw, int bytes) +void blk_fill_rwbs(char *rwbs, int op, u32 rw, int bytes) { int i = 0; if (rw & REQ_FLUSH) rwbs[i++] = 'F'; - if (rw & WRITE) + switch (op) { + case REQ_OP_WRITE: + case REQ_OP_WRITE_SAME: rwbs[i++] = 'W'; - else if (rw & REQ_DISCARD) + break; + case REQ_OP_DISCARD: rwbs[i++] = 'D'; - else if (bytes) + break; + case REQ_OP_READ: rwbs[i++] = 'R'; - else + break; + default: rwbs[i++] = 'N'; + } if (rw & REQ_FUA) rwbs[i++] = 'F'; -- cgit v1.2.3 From 3a5e02ced11e22ecd9da3d6710afe15bcfee1d10 Mon Sep 17 00:00:00 2001 From: Mike Christie Date: Sun, 5 Jun 2016 14:32:23 -0500 Subject: block, drivers: add REQ_OP_FLUSH operation This adds a REQ_OP_FLUSH operation that is sent to request_fn based drivers by the block layer's flush code, instead of sending requests with the request->cmd_flags REQ_FLUSH bit set. Signed-off-by: Mike Christie Reviewed-by: Christoph Hellwig Reviewed-by: Hannes Reinecke Signed-off-by: Jens Axboe --- kernel/trace/blktrace.c | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'kernel/trace') diff --git a/kernel/trace/blktrace.c b/kernel/trace/blktrace.c index 2d16fad519b2..0c70fbb6ea8d 100644 --- a/kernel/trace/blktrace.c +++ b/kernel/trace/blktrace.c @@ -223,6 +223,8 @@ static void __blk_add_trace(struct blk_trace *bt, sector_t sector, int bytes, what |= MASK_TC_BIT(op_flags, FUA); if (op == REQ_OP_DISCARD) what |= BLK_TC_ACT(BLK_TC_DISCARD); + if (op == REQ_OP_FLUSH) + what |= BLK_TC_ACT(BLK_TC_FLUSH); pid = tsk->pid; if (act_log_check(bt, what, sector, pid)) @@ -1788,6 +1790,9 @@ void blk_fill_rwbs(char *rwbs, int op, u32 rw, int bytes) case REQ_OP_DISCARD: rwbs[i++] = 'D'; break; + case REQ_OP_FLUSH: + rwbs[i++] = 'F'; + break; case REQ_OP_READ: rwbs[i++] = 'R'; break; -- cgit v1.2.3 From 28a8f0d317bf225ff15008f5dd66ae16242dd843 Mon Sep 17 00:00:00 2001 From: Mike Christie Date: Sun, 5 Jun 2016 14:32:25 -0500 Subject: block, drivers, fs: rename REQ_FLUSH to REQ_PREFLUSH To avoid confusion between REQ_OP_FLUSH, which is handled by request_fn drivers, and upper layers requesting the block layer perform a flush sequence along with possibly a WRITE, this patch renames REQ_FLUSH to REQ_PREFLUSH. Signed-off-by: Mike Christie Reviewed-by: Christoph Hellwig Reviewed-by: Hannes Reinecke Signed-off-by: Jens Axboe --- kernel/trace/blktrace.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/blktrace.c b/kernel/trace/blktrace.c index 0c70fbb6ea8d..03b0dd98ff0e 100644 --- a/kernel/trace/blktrace.c +++ b/kernel/trace/blktrace.c @@ -189,6 +189,7 @@ static const u32 ddir_act[2] = { BLK_TC_ACT(BLK_TC_READ), BLK_TC_ACT(BLK_TC_WRITE) }; #define BLK_TC_RAHEAD BLK_TC_AHEAD +#define BLK_TC_PREFLUSH BLK_TC_FLUSH /* The ilog2() calls fall out because they're constant */ #define MASK_TC_BIT(rw, __name) ((rw & REQ_ ## __name) << \ @@ -219,7 +220,7 @@ static void __blk_add_trace(struct blk_trace *bt, sector_t sector, int bytes, what |= MASK_TC_BIT(op_flags, SYNC); what |= MASK_TC_BIT(op_flags, RAHEAD); what |= MASK_TC_BIT(op_flags, META); - what |= MASK_TC_BIT(op_flags, FLUSH); + what |= MASK_TC_BIT(op_flags, PREFLUSH); what |= MASK_TC_BIT(op_flags, FUA); if (op == REQ_OP_DISCARD) what |= BLK_TC_ACT(BLK_TC_DISCARD); @@ -1779,7 +1780,7 @@ void blk_fill_rwbs(char *rwbs, int op, u32 rw, int bytes) { int i = 0; - if (rw & REQ_FLUSH) + if (rw & REQ_PREFLUSH) rwbs[i++] = 'F'; switch (op) { -- cgit v1.2.3 From 59a37f8baeb2c9d97f316584c90892d18bf846d4 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Fri, 17 Jun 2016 16:58:26 +0200 Subject: blktrace: avoid using timespec The blktrace code stores the current time in a 32-bit word in its user interface. This is a bad idea because 32-bit seconds overflow at some point. We probably have until 2106 before this one overflows, as it seems to use an 'unsigned' variable, but we should confirm that user space treats it the same way. Aside from this, we want to stop using 'struct timespec' here, so I'm adding a comment about the overflow and change the code to use timespec64 instead to make the loss of range more obvious. Signed-off-by: Arnd Bergmann Signed-off-by: Jens Axboe --- kernel/trace/blktrace.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/blktrace.c b/kernel/trace/blktrace.c index 03b0dd98ff0e..bedb84d168d1 100644 --- a/kernel/trace/blktrace.c +++ b/kernel/trace/blktrace.c @@ -127,12 +127,13 @@ static void trace_note_tsk(struct task_struct *tsk) static void trace_note_time(struct blk_trace *bt) { - struct timespec now; + struct timespec64 now; unsigned long flags; u32 words[2]; - getnstimeofday(&now); - words[0] = now.tv_sec; + /* need to check user space to see if this breaks in y2038 or y2106 */ + ktime_get_real_ts64(&now); + words[0] = (u32)now.tv_sec; words[1] = now.tv_nsec; local_irq_save(flags); -- cgit v1.2.3