From mboxrd@z Thu Jan 1 00:00:00 1970 From: huawei.libin@huawei.com (libin) Date: Mon, 30 Nov 2015 10:03:26 +0800 Subject: [PATCH] arm64: ftrace: stop using kstop_machine to enable/disable tracing In-Reply-To: <20151128105819.3451ab3e@grimm.local.home> References: <1448697009-17211-1-git-send-email-huawei.libin@huawei.com> <20151128105819.3451ab3e@grimm.local.home> Message-ID: <565BAE6E.9060309@huawei.com> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org on 2015/11/28 23:58, Steven Rostedt wrote: > On Sat, 28 Nov 2015 15:50:09 +0800 > Li Bin wrote: > >> On arm64, kstop_machine which is hugely disruptive to a running >> system is not needed to convert nops to ftrace calls or back, >> because that modifed code is a single 32bit instructions which >> is impossible to cross cache (or page) boundaries, and the used str >> instruction is single-copy atomic. > Is this really true? I thought that arm (and then perhaps arm64) has > some 2 byte instructions. If that's the case it is very well possible > that a 4 byte instruction can cross cache lines. When system in aarch32 state, it will use A32 or T32 instrucion set, and T32 (thumb) have 16-bit instructions. But arm64 that in aarch64 state only using A64 instruction set, which is a clean and fixed length instruction set that instuctions are always 32 bits wide. Right? Thanks, Li Bin > -- Steve > >> Cc: # 3.18+ >> Signed-off-by: Li Bin >> --- >> arch/arm64/kernel/ftrace.c | 5 +++++ >> 1 files changed, 5 insertions(+), 0 deletions(-) > . >