Date: Fri, 27 Mar 2015 13:46:30 +0000 From: bugzilla-noreply@freebsd.org To: freebsd-ports-bugs@FreeBSD.org Subject: [Bug 198920] [PATCH] www/squid: src/ipc/mem/Segment.cc patch to prevent memory pages flushing Message-ID: <bug-198920-13-jDQ7Zamk3w@https.bugs.freebsd.org/bugzilla/> In-Reply-To: <bug-198920-13@https.bugs.freebsd.org/bugzilla/> References: <bug-198920-13@https.bugs.freebsd.org/bugzilla/>
next in thread | previous in thread | raw e-mail | index | archive | help
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=198920 --- Comment #3 from emz@norma.perm.ru --- MAP_NOSYNC is FreeBSD-specific, Linux syscall simply doesn't have this flag. I used the "regression" term because without this flag PostgreSQL was terribly slow. On the other hand, squid isn't a database engine, it has lesser i/o density, but I though - if it still does use the mmap() call - why not. I didn't ask any of the squid developers, but I think I can ask Amos Jeffries, one of the main squid developers. But the main idea was "MAP_NOSYNC definitely cannot slow down squid, in fact, it's supposed to boost it up, so why not". I will update this bug with whatever Amos will reply. -- You are receiving this mail because: You are the assignee for the bug.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?bug-198920-13-jDQ7Zamk3w>