Date: Wed, 13 May 2015 14:56:01 +0000 From: "imp (Warner Losh)" <phabric-noreply@FreeBSD.org> To: freebsd-arm@freebsd.org Subject: [Differential] [Updated] D2377: Introduce ARM GICv3 support Message-ID: <7901b00e10c98e290f96cb7294082458@localhost.localdomain> In-Reply-To: <differential-rev-PHID-DREV-piql4a7htwxjub743777-req@FreeBSD.org> References: <differential-rev-PHID-DREV-piql4a7htwxjub743777-req@FreeBSD.org>
next in thread | previous in thread | raw e-mail | index | archive | help
imp added a comment. Two minor comments. One is a suggestion we might want to say something. The second is the 1s timeout seems too long and is badly implemented (DELAY(1) isn't the highest precision thing). I'm wondering why you'd ever need more than several milliseconds. Other than that, it looks OK. A couple of places could use some minor cleanup, but they are much better after the initial rounds of cleanup. INLINE COMMENTS sys/arm64/arm64/gic_v3.c:205 You should at least printf or panic or something :) sys/arm64/arm64/gic_v3.c:321 Does delay 1 really delay 1us every time over 1M iterations? I guess it isn't critical, but if you only want to wait a second for something time sources are more reliable. Then again, why wait a full second here? sys/arm64/arm64/gic_v3.c:535 Same comments here. why 1 second? REVISION DETAIL https://reviews.freebsd.org/D2377 EMAIL PREFERENCES https://reviews.freebsd.org/settings/panel/emailpreferences/ To: zbb, ian, emaste, andrew, imp, brueffer, joel, wblock Cc: kostikbel, emaste, andrew, imp, freebsd-arm
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?7901b00e10c98e290f96cb7294082458>