Kevin Oberman <[email protected]> writes: > Any hope for mozjemalloc? At this time I need to shutdown and restart > Firefox often and might be as bad as Chromium for memory use. > > i know that the malloc issue is at least a significant issue in memory use. > Is there a real intractable issue with getting the mozjemalloc working on > FreeBSD or is it just a matter of developer time?
"developer time" presumes there're developers. I'm just an #ifdef monkey. Porting a jemalloc fork requires understanding how jemalloc bootstraps. Old jemalloc (from FreeBSD 7-9) used spinlocks, no clue about the new one. mozjemalloc complicates this by code churn each Firefox release. To my knowledge, no one is working on porting mozjemalloc, so feel free to send patches. I can only help with build glue. > While I'm at it, what about OSS? Not nearly as critical, but would be nice. OSS in its unmaintained state is there for years as files/patch-bug1021761. pkg-message nowadays documents how to switch audio backend. Firefox build is simply broken without either ALSA, PULSEAUDIO or SNDIO, so you can't have pure OSS setup. To my knowledge, no one is working on finshing OSS support, so feel free to send patches. I can only help with build glue. _______________________________________________ [email protected] mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-gecko To unsubscribe, send any mail to "[email protected]"
