From mboxrd@z Thu Jan 1 00:00:00 1970 From: yingjoe.chen@mediatek.com (Yingjoe Chen) Date: Wed, 10 Dec 2014 17:19:26 +0800 Subject: [PATCH v5] clocksource: arch_timer: Fix code to use physical timers when requested In-Reply-To: <20141209165807.GJ31129@e104818-lin.cambridge.arm.com> References: <1416812564-26465-1-git-send-email-sonnyrao@chromium.org> <1417776064.14380.17.camel@mtksdaap41> <20141208162140.GN16185@e104818-lin.cambridge.arm.com> <1418106718.32622.31.camel@mtksdaap41> <20141209165807.GJ31129@e104818-lin.cambridge.arm.com> Message-ID: <1418203166.21309.3.camel@mtksdaap41> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org On Tue, 2014-12-09 at 16:58 +0000, Catalin Marinas wrote: <...> > You mix timer and counter terms here. Anyway, of we use physical timer, > you could argue that it makes sense to use the corresponding physical > counter (CNTPCT). However, on arm64 we only expose CNTVCT to user VDSO > and we want to use the same in the kernel. When booting at EL2, CNTVCT > == CNTPCT because we control CNTVOFF, that's unless we have some broken > firmware that does not restore CNTVOFF correctly. That's what we want > to spot early, hence the aim to always use the virtual counter (but not > the timer, use use the physical timer as it makes it easier for KVM). > > So the patch below, on top of linux-next, should solve the BUG(): Thanks for detail explanation and the patch. I tested it on my platform and it did solve the issue. So, Tested-by: Yingjoe Chen Joe.C