d/numam-spdk
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

test/bdevperf: add some more output and use LBA as data for verify

Browse Source 
When dumping results, include the LBA range used for the thread.
Also, to help with debug of verification erros, no longer use
rand data but instead use the current LBA.  This can provide
a clue on data mismatch as to what might be wrong.

Signed-off-by: paul luse <paul.e.luse@intel.com>
Change-Id: Ia02c35cf61ff2763c4a241314766e0bc65d18886
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/5177
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
Reviewed-by: <dongx.yi@intel.com>
This commit is contained in:
paul luse 2020-11-19 21:04:52 +00:00 committed by Tomasz Zawadzki
parent 7056bdf6ed
commit d80b4f4287
1 changed files with 17 additions and 10 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

27
test/bdev/bdevperf/bdevperf.c
View File

@ -227,8 +227,11 @@ performance_dump_job(struct bdevperf_aggregate_stats *stats, struct bdevperf_job
double io_per_second, mb_per_second, failed_per_second, timeout_per_second;
printf("\r Thread name: %s\n", spdk_thread_get_name(job->thread));
printf("\r Core Mask: 0x%s\n", spdk_cpuset_fmt(spdk_thread_get_cpumask(job->thread)));
printf("\t Core Mask: 0x%s\n", spdk_cpuset_fmt(spdk_thread_get_cpumask(job->thread)));
if (job->verify) {
printf("\t Verification LBA range: start 0x%lx length 0x%lx\n",
job->ios_base, job->size_in_ios);
}
if (stats->ema_period == 0) {
io_per_second = get_cma_io_per_second(job, stats->io_time_in_usec);
} else {
@ -238,10 +241,10 @@ performance_dump_job(struct bdevperf_aggregate_stats *stats, struct bdevperf_job
failed_per_second = (double)job->io_failed * 1000000 / stats->io_time_in_usec;
timeout_per_second = (double)job->io_timeout * 1000000 / stats->io_time_in_usec;
printf("\r %-20s: %10.2f IOPS %10.2f MiB/s\n",
printf("\t %-20s: %10.2f IOPS %10.2f MiB/s\n",
job->name, io_per_second, mb_per_second);
if (failed_per_second != 0) {
printf("\r %-20s: %10.2f Fail/s %8.2f TO/s\n",
printf("\t %-20s: %10.2f Fail/s %8.2f TO/s\n",
"", failed_per_second, timeout_per_second);
}
stats->total_io_per_second += io_per_second;
@ -252,9 +255,9 @@ performance_dump_job(struct bdevperf_aggregate_stats *stats, struct bdevperf_job
static void
generate_data(void *buf, int buf_len, int block_size, void *md_buf, int md_size,
int num_blocks, int seed)
int num_blocks)
{
int offset_blocks = 0, md_offset, data_block_size;
int offset_blocks = 0, md_offset, data_block_size, inner_offset;
if (buf_len < num_blocks * block_size) {
return;
@ -270,9 +273,13 @@ generate_data(void *buf, int buf_len, int block_size, void *md_buf, int md_size,
}
while (offset_blocks < num_blocks) {
memset(buf, seed, data_block_size);
memset(md_buf, seed, md_size);
buf += block_size;
inner_offset = 0;
while (inner_offset < data_block_size) {
*(uint32_t *)buf = offset_blocks;
inner_offset += sizeof(uint32_t);
buf += sizeof(uint32_t);
}
memset(md_buf, offset_blocks, md_size);
md_buf += md_offset;
offset_blocks++;
}
@ -861,7 +868,7 @@ bdevperf_submit_single(struct bdevperf_job *job, struct bdevperf_task *task)
generate_data(task->buf, job->buf_size,
spdk_bdev_get_block_size(job->bdev),
task->md_buf, spdk_bdev_get_md_size(job->bdev),
job->io_size_blocks, rand_r(&seed) % 256);
job->io_size_blocks);
if (g_zcopy) {
bdevperf_prep_zcopy_write_task(task);
return;