Date: Sun, 22 Feb 2015 09:20:52 +0000 From: "meloun-miracle-cz (Michal Meloun)" <phabric-noreply@FreeBSD.org> To: freebsd-arm@freebsd.org Subject: [Differential] [Commented On] D1833: Add memory barriers to buf_ring Message-ID: <460482fd8d4c5fc63ada8a2bb63b2e36@localhost.localdomain> In-Reply-To: <differential-rev-PHID-DREV-ntnotl4wj5wcj2eoukly-req@FreeBSD.org> References: <differential-rev-PHID-DREV-ntnotl4wj5wcj2eoukly-req@FreeBSD.org>
next in thread | previous in thread | raw e-mail | index | archive | help
meloun-miracle-cz added a comment. First of all, I was wrong in the behavior of atomic operations on ARM. After yesterday’s long session on IRC with Ian and Andy, we concluded that pure (non atomic) store affects(clears) state of exclusive monitor. So our implementation of atomics is fine on ARM. I apologize for my mistake, and, please, ignore everything about atomic_cmpset() in my previous message. Back to your first question. I thought it only as a technical note (because misordered read occurs on br_ring[]). So, i'm fine with proposed patch. And again, sorry for noise. I certainly did not intend to be “destructive”. REVISION DETAIL https://reviews.freebsd.org/D1833 To: zbb, kmacy, rpaulo, imp Cc: meloun-miracle-cz, onwahe-gmail-com, andrew, ian, adrian, freebsd-arm
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?460482fd8d4c5fc63ada8a2bb63b2e36>