And env vars up to 40%? Please fucking kill me.
-
-
coz-profiler might stop working on Rust in the near-future :(https://github.com/rust-lang/measureme/issues/39 …
1 reply 0 retweets 3 likes -
-
This reminds me: iirc Solaris has known tracepoint labels to mark mutexes. Don't believe Linux has something similar, but perhaps worth double checking?
1 reply 0 retweets 3 likes -
Replying to @yoshuawuyts @mgattozzi and
In its entirety I wish the stdlib had a stable tracepoint interface for consumption by exactly these kinds of tools, but that's probably a significant amount of work.
1 reply 0 retweets 3 likes -
I would kill for this, but like you said, lots of work
1 reply 0 retweets 3 likes -
Found the mutex probe docs for Solaris https://docs.oracle.com/cd/E36784_01/html/E36846/gkyeu.html … (Yeah sorry about the website)
1 reply 0 retweets 3 likes -
Look the most cartoonishly evil corp (they wear it like a badge of honor) bought em. Not much we can do about that.
1 reply 0 retweets 2 likes -
Heh yeah. (also replied to the thread suggesting we may want to introduce static tracepoints to restore Mutex probing -- not sure about the details, but figured it's worth bringing up)
1 reply 0 retweets 2 likes -
Replying to @yoshuawuyts @mgattozzi and
with eBPF you can trace pthread_mutex_(un)lock, that’s what http://deadlock.py from the bcc tools does: https://github.com/iovisor/bcc/blob/9e0817a9da/tools/deadlock.py …
1 reply 0 retweets 4 likes
Ohhhh, it even supports custom lock/unlock symbols! $ deadlock 181 --lock-symbols my_mutex_lock1 \ --unlock-symbols my_mutex_unlock1
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.