From mboxrd@z Thu Jan 1 00:00:00 1970 From: Christoffer Dall Subject: [PATCH v4 03/20] arm64: Use physical counter for in-kernel reads when booted in EL2 Date: Fri, 20 Oct 2017 13:49:22 +0200 Message-ID: <20171020114939.12554-4-christoffer.dall@linaro.org> References: <20171020114939.12554-1-christoffer.dall@linaro.org> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Cc: Christoffer Dall , kvm@vger.kernel.org, Marc Zyngier , Catalin Marinas , Will Deacon , Shih-Wei Li To: kvmarm@lists.cs.columbia.edu, linux-arm-kernel@lists.infradead.org Return-path: In-Reply-To: <20171020114939.12554-1-christoffer.dall@linaro.org> List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: kvmarm-bounces@lists.cs.columbia.edu Sender: kvmarm-bounces@lists.cs.columbia.edu List-Id: kvm.vger.kernel.org From: Christoffer Dall Using the physical counter allows KVM to retain the offset between the virtual and physical counter as long as it is actively running a VCPU. As soon as a VCPU is released, another thread is scheduled or we start running userspace applications, we reset the offset to 0, so that userspace accessing the virtual timer can still read the virtual counter and get the same view of time as the kernel. This opens up potential improvements for KVM performance, but we have to make a few adjustments to preserve system consistency. Currently get_cycles() is hardwired to arch_counter_get_cntvct() on arm64, but as we move to using the physical timer for the in-kernel time-keeping on systems that boot in EL2, we should use the same counter for get_cycles() as for other in-kernel timekeeping operations. Similarly, implementations of arch_timer_set_next_event_phys() is modified to use the counter specific to the timer being programmed. VHE kernels or kernels continuing to use the virtual timer are unaffected. Cc: Catalin Marinas Cc: Will Deacon Cc: Mark Rutland Cc: Marc Zyngier Signed-off-by: Christoffer Dall --- Notes: Changes since v3: - Consolidated the change of the arch_timer_read_counter function pointer and the change to get_cycles(). A new patch was added prior to this one to support stable reads of the physical counter including errata workaround. - Added !is_hyp_mode_available() clause to ARM64 systems when selecting the in-kernel counter, so that arm64 kernels not booting in EL2 remain unmodified. - Added EXPORT_SYMBOL_GPL(arch_timer_read_counter) to support building modules that reference get_cycles() and thereby arch_timer_read_counter. arch/arm64/include/asm/timex.h | 2 +- drivers/clocksource/arm_arch_timer.c | 12 ++++++++---- 2 files changed, 9 insertions(+), 5 deletions(-) diff --git a/arch/arm64/include/asm/timex.h b/arch/arm64/include/asm/timex.h index 81a076eb37fa..9ad60bae5c8d 100644 --- a/arch/arm64/include/asm/timex.h +++ b/arch/arm64/include/asm/timex.h @@ -22,7 +22,7 @@ * Use the current timer as a cycle counter since this is what we use for * the delay loop. */ -#define get_cycles() arch_counter_get_cntvct() +#define get_cycles() arch_timer_read_counter() #include diff --git a/drivers/clocksource/arm_arch_timer.c b/drivers/clocksource/arm_arch_timer.c index ff8f8a177156..061476e92db7 100644 --- a/drivers/clocksource/arm_arch_timer.c +++ b/drivers/clocksource/arm_arch_timer.c @@ -158,6 +158,7 @@ u32 arch_timer_reg_read(int access, enum arch_timer_reg reg, * if we don't have the cp15 accessors we won't have a problem. */ u64 (*arch_timer_read_counter)(void) = arch_counter_get_cntvct; +EXPORT_SYMBOL_GPL(arch_timer_read_counter); static u64 arch_counter_read(struct clocksource *cs) { @@ -329,16 +330,19 @@ static void erratum_set_next_event_tval_generic(const int access, unsigned long struct clock_event_device *clk) { unsigned long ctrl; - u64 cval = evt + arch_counter_get_cntvct(); + u64 cval; ctrl = arch_timer_reg_read(access, ARCH_TIMER_REG_CTRL, clk); ctrl |= ARCH_TIMER_CTRL_ENABLE; ctrl &= ~ARCH_TIMER_CTRL_IT_MASK; - if (access == ARCH_TIMER_PHYS_ACCESS) + if (access == ARCH_TIMER_PHYS_ACCESS) { + cval = evt + arch_counter_get_cntpct(); write_sysreg(cval, cntp_cval_el0); - else + } else { + cval = evt + arch_counter_get_cntvct(); write_sysreg(cval, cntv_cval_el0); + } arch_timer_reg_write(access, ARCH_TIMER_REG_CTRL, ctrl, clk); } @@ -913,7 +917,7 @@ static void __init arch_counter_register(unsigned type) /* Register the CP15 based counter if we have one */ if (type & ARCH_TIMER_TYPE_CP15) { - if (IS_ENABLED(CONFIG_ARM64) || + if ((IS_ENABLED(CONFIG_ARM64) && !is_hyp_mode_available()) || arch_timer_uses_ppi == ARCH_TIMER_VIRT_PPI) arch_timer_read_counter = arch_counter_get_cntvct; else -- 2.14.2 From mboxrd@z Thu Jan 1 00:00:00 1970 From: christoffer.dall@linaro.org (Christoffer Dall) Date: Fri, 20 Oct 2017 13:49:22 +0200 Subject: [PATCH v4 03/20] arm64: Use physical counter for in-kernel reads when booted in EL2 In-Reply-To: <20171020114939.12554-1-christoffer.dall@linaro.org> References: <20171020114939.12554-1-christoffer.dall@linaro.org> Message-ID: <20171020114939.12554-4-christoffer.dall@linaro.org> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org From: Christoffer Dall Using the physical counter allows KVM to retain the offset between the virtual and physical counter as long as it is actively running a VCPU. As soon as a VCPU is released, another thread is scheduled or we start running userspace applications, we reset the offset to 0, so that userspace accessing the virtual timer can still read the virtual counter and get the same view of time as the kernel. This opens up potential improvements for KVM performance, but we have to make a few adjustments to preserve system consistency. Currently get_cycles() is hardwired to arch_counter_get_cntvct() on arm64, but as we move to using the physical timer for the in-kernel time-keeping on systems that boot in EL2, we should use the same counter for get_cycles() as for other in-kernel timekeeping operations. Similarly, implementations of arch_timer_set_next_event_phys() is modified to use the counter specific to the timer being programmed. VHE kernels or kernels continuing to use the virtual timer are unaffected. Cc: Catalin Marinas Cc: Will Deacon Cc: Mark Rutland Cc: Marc Zyngier Signed-off-by: Christoffer Dall --- Notes: Changes since v3: - Consolidated the change of the arch_timer_read_counter function pointer and the change to get_cycles(). A new patch was added prior to this one to support stable reads of the physical counter including errata workaround. - Added !is_hyp_mode_available() clause to ARM64 systems when selecting the in-kernel counter, so that arm64 kernels not booting in EL2 remain unmodified. - Added EXPORT_SYMBOL_GPL(arch_timer_read_counter) to support building modules that reference get_cycles() and thereby arch_timer_read_counter. arch/arm64/include/asm/timex.h | 2 +- drivers/clocksource/arm_arch_timer.c | 12 ++++++++---- 2 files changed, 9 insertions(+), 5 deletions(-) diff --git a/arch/arm64/include/asm/timex.h b/arch/arm64/include/asm/timex.h index 81a076eb37fa..9ad60bae5c8d 100644 --- a/arch/arm64/include/asm/timex.h +++ b/arch/arm64/include/asm/timex.h @@ -22,7 +22,7 @@ * Use the current timer as a cycle counter since this is what we use for * the delay loop. */ -#define get_cycles() arch_counter_get_cntvct() +#define get_cycles() arch_timer_read_counter() #include diff --git a/drivers/clocksource/arm_arch_timer.c b/drivers/clocksource/arm_arch_timer.c index ff8f8a177156..061476e92db7 100644 --- a/drivers/clocksource/arm_arch_timer.c +++ b/drivers/clocksource/arm_arch_timer.c @@ -158,6 +158,7 @@ u32 arch_timer_reg_read(int access, enum arch_timer_reg reg, * if we don't have the cp15 accessors we won't have a problem. */ u64 (*arch_timer_read_counter)(void) = arch_counter_get_cntvct; +EXPORT_SYMBOL_GPL(arch_timer_read_counter); static u64 arch_counter_read(struct clocksource *cs) { @@ -329,16 +330,19 @@ static void erratum_set_next_event_tval_generic(const int access, unsigned long struct clock_event_device *clk) { unsigned long ctrl; - u64 cval = evt + arch_counter_get_cntvct(); + u64 cval; ctrl = arch_timer_reg_read(access, ARCH_TIMER_REG_CTRL, clk); ctrl |= ARCH_TIMER_CTRL_ENABLE; ctrl &= ~ARCH_TIMER_CTRL_IT_MASK; - if (access == ARCH_TIMER_PHYS_ACCESS) + if (access == ARCH_TIMER_PHYS_ACCESS) { + cval = evt + arch_counter_get_cntpct(); write_sysreg(cval, cntp_cval_el0); - else + } else { + cval = evt + arch_counter_get_cntvct(); write_sysreg(cval, cntv_cval_el0); + } arch_timer_reg_write(access, ARCH_TIMER_REG_CTRL, ctrl, clk); } @@ -913,7 +917,7 @@ static void __init arch_counter_register(unsigned type) /* Register the CP15 based counter if we have one */ if (type & ARCH_TIMER_TYPE_CP15) { - if (IS_ENABLED(CONFIG_ARM64) || + if ((IS_ENABLED(CONFIG_ARM64) && !is_hyp_mode_available()) || arch_timer_uses_ppi == ARCH_TIMER_VIRT_PPI) arch_timer_read_counter = arch_counter_get_cntvct; else -- 2.14.2