• Joined on 2025年07月26日
oech3 closed issue ivarch/pv#200 2026年06月08日 01:51:30 +02:00
strange throuput change of uutils/yes between 1.10.5~main
oech3 commented on issue ivarch/pv#188 2026年06月05日 19:01:10 +02:00
pv<sparse is slower than cat sparse | pv with cat supporting splice

I don't know why, and I still don't believe it, but splice with middler pipe is faster than direct splice. See https://github.com/uutils/coreutils/pull/12651#issue-4598546707 .

oech3 commented on issue ivarch/pv#188 2026年06月05日 11:07:14 +02:00
pv<sparse is slower than cat sparse | pv with cat supporting splice

Not sure. Did this change broke measurement accuracy if inpit is coming from splice? ivarch/pv#200

oech3 opened issue ivarch/pv#200 2026年06月05日 10:55:40 +02:00
strange throuput change of uutils/yes between 1.10.5~main
oech3 commented on issue ivarch/pv#199 2026年06月04日 06:00:49 +02:00
remove stat

The most reliable ways to check that we can bypass middler pipe, and use copy_file_range are just trying direct splice and copy_file_range for given fd's.

oech3 commented on issue ivarch/pv#196 2026年06月03日 06:20:10 +02:00
Use copy_file_range(2) where available
oech3 commented on issue ivarch/pv#192 2026年06月03日 02:22:19 +02:00
unshare -rm bash -c "mount --bind /proc/self/mem /dev/null && pv <huge -X" pv: /dev/null: No such process

Closing since this appears between main and latest stable (thus, people no longer see this). Thankyou for e74cab6985 .

oech3 closed issue ivarch/pv#192 2026年06月03日 02:20:21 +02:00
unshare -rm bash -c "mount --bind /proc/self/mem /dev/null && pv <huge -X" pv: /dev/null: No such process
oech3 commented on issue ivarch/pv#196 2026年06月03日 02:10:14 +02:00
Use copy_file_range(2) where available

Unfortunately, --disable-copy-file-range improved performance on tmpfs to tmpfs (repeat and format is omitted):

oech3 commented on issue ivarch/pv#196 2026年06月03日 02:03:05 +02:00
Use copy_file_range(2) where available

I cannot see ---> at pv to tmpfs from btrfs. So it seems cross-device copy_file_range is not supported.

oech3 opened issue ivarch/pv#199 2026年06月03日 01:48:31 +02:00
remove stat
oech3 commented on issue ivarch/pv#196 2026年06月02日 17:39:27 +02:00
Use copy_file_range(2) where available

copy_file_range (or sendfile) internally splice's on non- CoW system. Can we change middler pipe size in the case? cross-device copy by splice might faster than copy_file_range.

oech3 commented on issue ivarch/pv#192 2026年05月26日 07:44:12 +02:00
unshare -rm bash -c "mount --bind /proc/self/mem /dev/null && pv <huge -X" pv: /dev/null: No such process

I think read fallback is enough (no one tries to use pv -X under container).

oech3 commented on issue ivarch/pv#192 2026年05月25日 17:05:52 +02:00
unshare -rm bash -c "mount --bind /proc/self/mem /dev/null && pv <huge -X" pv: /dev/null: No such process

I should benchmark splice to /dev/null vs splice to pipe & regenerate pipe with fcntl.

oech3 commented on issue ivarch/pv#181 2026年05月04日 16:52:23 +02:00
Option to produce a hash of the input

Is this for performance (with pv splice)? I have a pending PR on uutils/tee to support zero-copy and I guess GNU coreutils will do that soon.

oech3 commented on issue ivarch/pv#192 2026年05月04日 16:42:20 +02:00
unshare -rm bash -c "mount --bind /proc/self/mem /dev/null && pv <huge -X" pv: /dev/null: No such process

And exit code 0 is not suitable for scripts.

oech3 commented on issue ivarch/pv#192 2026年05月03日 19:41:16 +02:00
unshare -rm bash -c "mount --bind /proc/self/mem /dev/null && pv <huge -X" pv: /dev/null: No such process

I considered that /dev/null is useful only on Linux since splice is Linux specific.

oech3 opened issue ivarch/pv#192 2026年05月03日 19:17:52 +02:00
unshare -rm bash -c "mount --bind /proc/self/mem /dev/null && pv <huge -X" pv: /dev/null: No such process
oech3 closed issue ivarch/pv#191 2026年05月03日 19:12:44 +02:00
-X, --discard is slow if input is splice(2)'d