From mboxrd@z Thu Jan 1 00:00:00 1970 From: Catalin Marinas Subject: Re: [PATCH v3 21/26] arm64: Introduce asm/vdso/arch_timer.h Date: Sun, 15 Mar 2020 18:32:16 +0000 Message-ID: <20200315183151.GE32205@mbp> References: <20200313154345.56760-1-vincenzo.frascino@arm.com> <20200313154345.56760-22-vincenzo.frascino@arm.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Content-Disposition: inline In-Reply-To: <20200313154345.56760-22-vincenzo.frascino@arm.com> Sender: linux-mips-owner@vger.kernel.org To: Vincenzo Frascino Cc: linux-arch@vger.kernel.org, linux-arm-kernel@lists.infradead.org, linux-kernel@vger.kernel.org, linux-mips@vger.kernel.org, clang-built-linux@googlegroups.com, x86@kernel.org, Will Deacon , Arnd Bergmann , Russell King , Paul Burton , Thomas Gleixner , Andy Lutomirski , Ingo Molnar , Borislav Petkov , Stephen Boyd , Mark Salyzyn , Kees Cook , Peter Collingbourne , Dmitry Safonov <0x7f454c46@gmail.com>, Andrei Vagin , Nick Desaulniers , Marc Zyngier Mark List-Id: linux-arch.vger.kernel.org On Fri, Mar 13, 2020 at 03:43:40PM +0000, Vincenzo Frascino wrote: > The vDSO library should only include the necessary headers required for > a userspace library (UAPI and a minimal set of kernel headers). To make > this possible it is necessary to isolate from the kernel headers the > common parts that are strictly necessary to build the library. > > Introduce asm/vdso/arch_timer.h to contain all the arm64 specific > code. This allows to replace the second isb() in __arch_get_hw_counter() > with a fake dependent stack read of the counter which improves the vdso > library peformances of ~4.5%. Below the results of vdsotest [1] ran for > 100 iterations. The subject seems to imply a non-functional change but as you read, it gets a lot more complicated. Could you keep the functional change separate from the header clean-up, maybe submit it as an independent patch? And it shouldn't go in without Will's ack ;). -- Catalin From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from foss.arm.com ([217.140.110.172]:34626 "EHLO foss.arm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1727399AbgCOScb (ORCPT ); Sun, 15 Mar 2020 14:32:31 -0400 Date: Sun, 15 Mar 2020 18:32:16 +0000 From: Catalin Marinas Subject: Re: [PATCH v3 21/26] arm64: Introduce asm/vdso/arch_timer.h Message-ID: <20200315183151.GE32205@mbp> References: <20200313154345.56760-1-vincenzo.frascino@arm.com> <20200313154345.56760-22-vincenzo.frascino@arm.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20200313154345.56760-22-vincenzo.frascino@arm.com> Sender: linux-arch-owner@vger.kernel.org List-ID: To: Vincenzo Frascino Cc: linux-arch@vger.kernel.org, linux-arm-kernel@lists.infradead.org, linux-kernel@vger.kernel.org, linux-mips@vger.kernel.org, clang-built-linux@googlegroups.com, x86@kernel.org, Will Deacon , Arnd Bergmann , Russell King , Paul Burton , Thomas Gleixner , Andy Lutomirski , Ingo Molnar , Borislav Petkov , Stephen Boyd , Mark Salyzyn , Kees Cook , Peter Collingbourne , Dmitry Safonov <0x7f454c46@gmail.com>, Andrei Vagin , Nick Desaulniers , Marc Zyngier , Mark Rutland , Will Deacon Message-ID: <20200315183216.2leg9tQYoFYk1-f6ZXfzQAd8hNKr-SDsDJ_fFJYTQec@z> On Fri, Mar 13, 2020 at 03:43:40PM +0000, Vincenzo Frascino wrote: > The vDSO library should only include the necessary headers required for > a userspace library (UAPI and a minimal set of kernel headers). To make > this possible it is necessary to isolate from the kernel headers the > common parts that are strictly necessary to build the library. > > Introduce asm/vdso/arch_timer.h to contain all the arm64 specific > code. This allows to replace the second isb() in __arch_get_hw_counter() > with a fake dependent stack read of the counter which improves the vdso > library peformances of ~4.5%. Below the results of vdsotest [1] ran for > 100 iterations. The subject seems to imply a non-functional change but as you read, it gets a lot more complicated. Could you keep the functional change separate from the header clean-up, maybe submit it as an independent patch? And it shouldn't go in without Will's ack ;). -- Catalin