On 12/04/2016 04:28 PM, Chris Murphy wrote:
4.8.11-300.fc25.x86_64 I'm currently doing a btrfs send/receive and I'm seeing a rather large hit for crc32c, bigger than aes-ni (the volume is on dm crypt), using perf top. 14.03% btrfs [.] __crc32c_le 10.50% [kernel] [k] _aesni_enc4
This is surprising, although send/recv does do a lot of small crc runs. What is the overall CPU usage? Maybe pin btrfs to a single CPU and use mpstat to see how hot that one CPU is. If we're 14% of a CPU running at 100%, that's a big deal. If we're 14% of a CPU running at 5%, we safely ignore it.
Complete output is here for 1 month: https://urldefense.proofpoint.com/v2/url?u=https-3A__paste.fedoraproject.org_498914_&d=DgIBaQ&c=5VD0RTtNlTh3ycd41b3MUw&r=9QPtTAxcitoznaWRKKHoEQ&m=YXgqrDMWJP2u2reRh04tnYz4CrO5SPUhSHd1cF2OcR0&s=axoqoaw0ZkYXdDxlTODLOfzqEK7uIJHqMsBjpkyIw5o&e= I don't remember crc32's taking this much CPU before, so it seems like a regression but offhand I don't know when it started. [chris@f25s ~]$ dmesg | grep crc32 [ 4.226700] Btrfs loaded, crc32c=crc32c-intel
At least we know you're using the intel accelerated one. Every time someone posts this dmesg output to the list, I owe Jeff another beer.
-chris -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
