* [PATCH v12 0/8] Clock for CPU scaling support for msm8996
From: Amit Kucheria @ 2018-05-24 11:25 UTC (permalink / raw)
To: linux-arm-kernel
In-Reply-To: <1527148218-16540-1-git-send-email-ilialin@codeaurora.org>
On Thu, May 24, 2018 at 10:50 AM, Ilia Lin <ilialin@codeaurora.org> wrote:
> [v12]
> * Addressed the kbuild fail on arm architecture
>
> [v11]
> * Split the series into domains
>
> [v9]
> * Addressed comments from Viresh and Russel about the error handling
>
> [v8]
> * Reordered the patch series into 4 groups
> * Addressed comments from Amit about the comments and commit messages
> * Addressed comments from Amit and Viresh about the resourses deallocation
>
> [v7]
> * Addressed comments from Viresh about resourses deallocation
> and DT compatible
>
> [v6]
> * Addressed comments from Viresh about:
> ** Comments style
> ** Kconfig bool instead of tristate
> ** DT and documentation style
> ** Resourses deallocation on an error
> ** Typos
>
> [v5]
> * Rebased
> * Addressed comments from Bjorn about SPDX style,
> functions and parameters naming
> * Addressed comments from Viresh DT properties and style, comments style,
> resourses deallocation, documentation placement
> * Addressed comments from Sricharan about unnessesary include
> * Addressed comments from Nicolas
> * Addressed comments from Rob about the commit messages and acks
> * Addressed comments from Mark
>
> [v4]
> * Adressed all comments from Stephen
> * Added CPU regulator support
> * Added qcom-cpufreq-kryo driver
>
> [v3]
> * Rebased on top of the latest PLL driver changes
> * Addressed comment from Rob Herring for bindings
>
> [v2]
> * Addressed comments from Rob Herring for bindings
> * Addressed comments from Mark Rutland for memory barrier
> * Addressed comments from Julien Thierry for clock reenabling condition
> * Tuned the HW configuration for clock frequencies below 600MHz
>
> SOC (1/15):
> Extracts the kryo l2 accessors driver from the QCOM PMU driver
>
> Clocks (2/15-9/15):
> This series adds support for the CPU clocks on msm8996 devices.
> The driver uses the existing PLL drivers and is required to control
> the CPU frequency scaling on the MSM8996.
>
> Ilia Lin (6):
> soc: qcom: Separate kryo l2 accessors from PMU driver
> clk: Use devm_ in the register fixed factor clock
> clk: qcom: Add CPU clock driver for msm8996
> dt-bindings: clk: qcom: Add bindings for CPU clock for msm8996
> clk: qcom: cpu-8996: Add support to switch below 600Mhz
> clk: qcom: Add ACD path to CPU clock driver for msm8996
>
> Rajendra Nayak (2):
> clk: qcom: Make clk_alpha_pll_configure available to modules
> clk: qcom: cpu-8996: Add support to switch to alternate PLL
>
> .../devicetree/bindings/clock/qcom,kryocc.txt | 17 +
> drivers/clk/clk-fixed-factor.c | 2 +-
> drivers/clk/qcom/Kconfig | 10 +
> drivers/clk/qcom/Makefile | 1 +
> drivers/clk/qcom/clk-alpha-pll.c | 1 +
> drivers/clk/qcom/clk-alpha-pll.h | 6 +
> drivers/clk/qcom/clk-cpu-8996.c | 510 +++++++++++++++++++++
> drivers/perf/Kconfig | 1 +
> drivers/perf/qcom_l2_pmu.c | 90 +---
> drivers/soc/qcom/Kconfig | 3 +
> drivers/soc/qcom/Makefile | 1 +
> drivers/soc/qcom/kryo-l2-accessors.c | 56 +++
> include/soc/qcom/kryo-l2-accessors.h | 12 +
For the perf/l2-accessors part, Reviewed-by: Amit Kucheria
<amit.kucheria@linaro.org>
For the entire series, Tested-by: Amit Kucheria <amit.kucheria@linaro.org>
^ permalink raw reply
* [PATCH 02/14] arm64: Call ARCH_WORKAROUND_2 on transitions between EL0 and EL1
From: Mark Rutland @ 2018-05-24 11:23 UTC (permalink / raw)
To: linux-arm-kernel
In-Reply-To: <20180524110058.mrkwc3rx5l3v6wad@lakrids.cambridge.arm.com>
On Thu, May 24, 2018 at 12:00:58PM +0100, Mark Rutland wrote:
> On Tue, May 22, 2018 at 04:06:36PM +0100, Marc Zyngier wrote:
> > In order for the kernel to protect itself, let's call the SSBD mitigation
> > implemented by the higher exception level (either hypervisor or firmware)
> > on each transition between userspace and kernel.
> >
> > We must take the PSCI conduit into account in order to target the
> > right exception level, hence the introduction of a runtime patching
> > callback.
> >
> > Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
> > ---
> > arch/arm64/kernel/cpu_errata.c | 18 ++++++++++++++++++
> > arch/arm64/kernel/entry.S | 22 ++++++++++++++++++++++
> > include/linux/arm-smccc.h | 5 +++++
> > 3 files changed, 45 insertions(+)
> >
> > diff --git a/arch/arm64/kernel/cpu_errata.c b/arch/arm64/kernel/cpu_errata.c
> > index a900befadfe8..46b3aafb631a 100644
> > --- a/arch/arm64/kernel/cpu_errata.c
> > +++ b/arch/arm64/kernel/cpu_errata.c
> > @@ -232,6 +232,24 @@ enable_smccc_arch_workaround_1(const struct arm64_cpu_capabilities *entry)
> > }
> > #endif /* CONFIG_HARDEN_BRANCH_PREDICTOR */
> >
> > +#ifdef CONFIG_ARM64_SSBD
> > +void __init arm64_update_smccc_conduit(struct alt_instr *alt,
> > + __le32 *origptr, __le32 *updptr,
> > + int nr_inst)
> > +{
> > + u32 insn;
> > +
> > + BUG_ON(nr_inst != 1);
> > +
> > + if (psci_ops.conduit == PSCI_CONDUIT_HVC)
> > + insn = aarch64_insn_get_hvc_value();
> > + else
> > + insn = aarch64_insn_get_smc_value();
>
> Shouldn't this also handle the case where there is no conduit?
Due to the config symbol not being defined yet, and various other fixups
in later patches, this is actually benign.
However, if you make this:
switch (psci_ops.conduit) {
case PSCI_CONDUIT_NONE:
return;
case PSCI_CONDUIT_HVC:
insn = aarch64_insn_get_hvc_value();
break;
case PSCI_CONDUIT_SMC:
insn = aarch64_insn_get_smc_value();
break;
}
... then we won't even bother patching the nop in the default case
regardless, which is nicer, IMO.
With that:
Reviewed-by: Mark Rutland <mark.rutland@arm.com>
Thanks,
Mark.
>
> See below comment in apply_ssbd for rationale.
>
> > +
> > + *updptr = cpu_to_le32(insn);
> > +}
> > +#endif /* CONFIG_ARM64_SSBD */
> > +
> > #define CAP_MIDR_RANGE(model, v_min, r_min, v_max, r_max) \
> > .matches = is_affected_midr_range, \
> > .midr_range = MIDR_RANGE(model, v_min, r_min, v_max, r_max)
> > diff --git a/arch/arm64/kernel/entry.S b/arch/arm64/kernel/entry.S
> > index ec2ee720e33e..f33e6aed3037 100644
> > --- a/arch/arm64/kernel/entry.S
> > +++ b/arch/arm64/kernel/entry.S
> > @@ -18,6 +18,7 @@
> > * along with this program. If not, see <http://www.gnu.org/licenses/>.
> > */
> >
> > +#include <linux/arm-smccc.h>
> > #include <linux/init.h>
> > #include <linux/linkage.h>
> >
> > @@ -137,6 +138,18 @@ alternative_else_nop_endif
> > add \dst, \dst, #(\sym - .entry.tramp.text)
> > .endm
> >
> > + // This macro corrupts x0-x3. It is the caller's duty
> > + // to save/restore them if required.
> > + .macro apply_ssbd, state
> > +#ifdef CONFIG_ARM64_SSBD
> > + mov w0, #ARM_SMCCC_ARCH_WORKAROUND_2
> > + mov w1, #\state
> > +alternative_cb arm64_update_smccc_conduit
> > + nop // Patched to SMC/HVC #0
> > +alternative_cb_end
> > +#endif
> > + .endm
>
> If my system doesn't have SMCCC1.1, or the FW doesn't have an
> implementation of ARCH_WORKAROUND_2, does this stay as a NOP?
>
> It looks like this would be patched to an SMC, which would be fatal on
> systems without EL3 FW.
>
> > +
> > .macro kernel_entry, el, regsize = 64
> > .if \regsize == 32
> > mov w0, w0 // zero upper 32 bits of x0
> > @@ -163,6 +176,13 @@ alternative_else_nop_endif
> > ldr x19, [tsk, #TSK_TI_FLAGS] // since we can unmask debug
> > disable_step_tsk x19, x20 // exceptions when scheduling.
> >
> > + apply_ssbd 1
>
>
> ... and thus kernel_entry would be fatal.
>
> Thanks,
> Mark.
^ permalink raw reply
* [PATCH v10 13/18] arm64/sve: Move sve_pffr() to fpsimd.h and make inline
From: Dave Martin @ 2018-05-24 11:22 UTC (permalink / raw)
To: linux-arm-kernel
In-Reply-To: <8736yhtkvo.fsf@linaro.org>
On Thu, May 24, 2018 at 11:20:59AM +0100, Alex Benn?e wrote:
>
> Dave Martin <Dave.Martin@arm.com> writes:
>
> > In order to make sve_save_state()/sve_load_state() more easily
> > reusable and to get rid of a potential branch on context switch
> > critical paths, this patch makes sve_pffr() inline and moves it to
> > fpsimd.h.
> >
> > <asm/processor.h> must be included in fpsimd.h in order to make
> > this work, and this creates an #include cycle that is tricky to
> > avoid without modifying core code, due to the way the PR_SVE_*()
> > prctl helpers are included in the core prctl implementation.
> >
> > Instead of breaking the cycle, this patch defers inclusion of
> > <asm/fpsimd.h> in <asm/processor.h> until the point where it is
> > actually needed: i.e., immediately before the prctl definitions.
> >
> > No functional change.
> >
> > Signed-off-by: Dave Martin <Dave.Martin@arm.com>
> > Acked-by: Catalin Marinas <catalin.marinas@arm.com>
> > Acked-by: Marc Zyngier <marc.zyngier@arm.com>
> > ---
> > arch/arm64/include/asm/fpsimd.h | 13 +++++++++++++
> > arch/arm64/include/asm/processor.h | 3 ++-
> > arch/arm64/kernel/fpsimd.c | 12 ------------
> > 3 files changed, 15 insertions(+), 13 deletions(-)
> >
> > diff --git a/arch/arm64/include/asm/fpsimd.h b/arch/arm64/include/asm/fpsimd.h
> > index fb60b22..fa92747 100644
> > --- a/arch/arm64/include/asm/fpsimd.h
> > +++ b/arch/arm64/include/asm/fpsimd.h
> > @@ -18,6 +18,8 @@
> >
> > #include <asm/ptrace.h>
> > #include <asm/errno.h>
> > +#include <asm/processor.h>
> > +#include <asm/sigcontext.h>
> >
> > #ifndef __ASSEMBLY__
> >
> > @@ -61,6 +63,17 @@ extern void sve_flush_cpu_state(void);
> > /* Maximum VL that SVE VL-agnostic software can transparently support */
> > #define SVE_VL_ARCH_MAX 0x100
> >
> > +/* Offset of FFR in the SVE register dump */
> > +static inline size_t sve_ffr_offset(int vl)
> > +{
> > + return SVE_SIG_FFR_OFFSET(sve_vq_from_vl(vl)) - SVE_SIG_REGS_OFFSET;
> > +}
> > +
> > +static inline void *sve_pffr(struct thread_struct *thread)
> > +{
> > + return (char *)thread->sve_state + sve_ffr_offset(thread->sve_vl);
> > +}
> > +
> > extern void sve_save_state(void *state, u32 *pfpsr);
> > extern void sve_load_state(void const *state, u32 const *pfpsr,
> > unsigned long vq_minus_1);
> > diff --git a/arch/arm64/include/asm/processor.h b/arch/arm64/include/asm/processor.h
> > index f902b6d..ebaadb1 100644
> > --- a/arch/arm64/include/asm/processor.h
> > +++ b/arch/arm64/include/asm/processor.h
> > @@ -40,7 +40,6 @@
> >
> > #include <asm/alternative.h>
> > #include <asm/cpufeature.h>
> > -#include <asm/fpsimd.h>
> > #include <asm/hw_breakpoint.h>
> > #include <asm/lse.h>
> > #include <asm/pgtable-hwdef.h>
> > @@ -245,6 +244,8 @@ void cpu_enable_pan(const struct arm64_cpu_capabilities *__unused);
> > void cpu_enable_cache_maint_trap(const struct arm64_cpu_capabilities *__unused);
> > void cpu_clear_disr(const struct arm64_cpu_capabilities *__unused);
> >
> > +#include <asm/fpsimd.h>
> > +
>
> You really need a one-liner comment to note why the include is in a
> funny place to save someone just moving it back and then getting really
> confused. Maybe:
>
> /* included just in time to avoid circular inclusion issues */
> #include <asm/fpsimd.h>
>
> It still seems weird to me though :-/
How about
/*
* Not at the top of the file due to a direct #include cycle between
* <asm/fpsimd.h> and <asm/processor.h>. Deferring this #include
* ensures that contents of processor.h are visible to fpsimd.h even if
* processor.h is included first.
*
* These prctl helpers are the only things in this file that require
* fpsimd.h. The core code expects them to be in this header.
*/
?
Cheers
---Dave
>
> Otherwise:
>
> Reviewed-by: Alex Benn?e <alex.bennee@linaro.org>
^ permalink raw reply
* [PATCH v4] arm64: allwinner: a64: Add Amarula A64-Relic initial support
From: Jagan Teki @ 2018-05-24 11:16 UTC (permalink / raw)
To: linux-arm-kernel
Amarula A64-Relic is Allwinner A64 based IoT device, which support
- Allwinner A64 Cortex-A53
- Mali-400MP2 GPU
- AXP803 PMIC
- 1GB DDR3 RAM
- 8GB eMMC
- AP6330 Wifi/BLE
- MIPI-DSI
- CSI: OV5640 sensor
- USB OTG
- 12V DC power supply
Signed-off-by: Jagan Teki <jagan@amarulasolutions.com>
---
Changes for v4:
- update dr_mode as 'otg'
Changes for v3:
- Use sun50i-a64-amarula-relic.dts name
- add eldo3 for dvdd-csi
- update dldo4 min voltage as 3.3v as per schematics
- use dldo3 name as dovdd-csi
- update aldo1, aldo2 voltages as per schematics
Changes for v2:
- Rename dts name to sun50i-a64-relic.dts which is simple to use
- Update dldo4 min voltage as 1.8
- Use licence year as 2018
arch/arm64/boot/dts/allwinner/Makefile | 1 +
.../dts/allwinner/sun50i-a64-amarula-relic.dts | 188 +++++++++++++++++++++
2 files changed, 189 insertions(+)
create mode 100644 arch/arm64/boot/dts/allwinner/sun50i-a64-amarula-relic.dts
diff --git a/arch/arm64/boot/dts/allwinner/Makefile b/arch/arm64/boot/dts/allwinner/Makefile
index c31f90a49481..67ce8c500b2e 100644
--- a/arch/arm64/boot/dts/allwinner/Makefile
+++ b/arch/arm64/boot/dts/allwinner/Makefile
@@ -1,4 +1,5 @@
# SPDX-License-Identifier: GPL-2.0
+dtb-$(CONFIG_ARCH_SUNXI) += sun50i-a64-amarula-relic.dtb
dtb-$(CONFIG_ARCH_SUNXI) += sun50i-a64-bananapi-m64.dtb
dtb-$(CONFIG_ARCH_SUNXI) += sun50i-a64-nanopi-a64.dtb
dtb-$(CONFIG_ARCH_SUNXI) += sun50i-a64-olinuxino.dtb
diff --git a/arch/arm64/boot/dts/allwinner/sun50i-a64-amarula-relic.dts b/arch/arm64/boot/dts/allwinner/sun50i-a64-amarula-relic.dts
new file mode 100644
index 000000000000..ce4a256ff086
--- /dev/null
+++ b/arch/arm64/boot/dts/allwinner/sun50i-a64-amarula-relic.dts
@@ -0,0 +1,188 @@
+// SPDX-License-Identifier: (GPL-2.0+ OR MIT)
+/*
+ * Copyright (C) 2018 Amarula Solutions B.V.
+ * Author: Jagan Teki <jagan@amarulasolutions.com>
+ */
+
+/dts-v1/;
+
+#include "sun50i-a64.dtsi"
+
+#include <dt-bindings/gpio/gpio.h>
+
+/ {
+ model = "Amarula A64-Relic";
+ compatible = "amarula,a64-relic", "allwinner,sun50i-a64";
+
+ aliases {
+ serial0 = &uart0;
+ };
+
+ chosen {
+ stdout-path = "serial0:115200n8";
+ };
+};
+
+&ehci0 {
+ status = "okay";
+};
+
+&mmc2 {
+ pinctrl-names = "default";
+ pinctrl-0 = <&mmc2_pins>;
+ vmmc-supply = <®_dcdc1>;
+ bus-width = <8>;
+ non-removable;
+ cap-mmc-hw-reset;
+ status = "okay";
+};
+
+&ohci0 {
+ status = "okay";
+};
+
+&r_rsb {
+ status = "okay";
+
+ axp803: pmic at 3a3 {
+ compatible = "x-powers,axp803";
+ reg = <0x3a3>;
+ interrupt-parent = <&r_intc>;
+ interrupts = <0 IRQ_TYPE_LEVEL_LOW>;
+ x-powers,drive-vbus-en; /* set N_VBUSEN as output pin */
+ };
+};
+
+#include "axp803.dtsi"
+
+®_aldo1 {
+ regulator-always-on;
+ regulator-min-microvolt = <2800000>;
+ regulator-max-microvolt = <2800000>;
+ regulator-name = "avdd-csi";
+};
+
+®_aldo2 {
+ regulator-always-on;
+ regulator-min-microvolt = <1800000>;
+ regulator-max-microvolt = <3300000>;
+ regulator-name = "vcc-pl";
+};
+
+®_aldo3 {
+ regulator-always-on;
+ regulator-min-microvolt = <3000000>;
+ regulator-max-microvolt = <3000000>;
+ regulator-name = "vcc-pll-avcc";
+};
+
+®_dcdc1 {
+ regulator-always-on;
+ regulator-min-microvolt = <3300000>;
+ regulator-max-microvolt = <3300000>;
+ regulator-name = "vcc-3v3";
+};
+
+®_dcdc2 {
+ regulator-always-on;
+ regulator-min-microvolt = <1040000>;
+ regulator-max-microvolt = <1300000>;
+ regulator-name = "vdd-cpux";
+};
+
+/* DCDC3 is polyphased with DCDC2 */
+
+®_dcdc5 {
+ regulator-always-on;
+ regulator-min-microvolt = <1500000>;
+ regulator-max-microvolt = <1500000>;
+ regulator-name = "vcc-dram";
+};
+
+®_dcdc6 {
+ regulator-always-on;
+ regulator-min-microvolt = <1100000>;
+ regulator-max-microvolt = <1100000>;
+ regulator-name = "vdd-sys";
+};
+
+®_dldo1 {
+ regulator-min-microvolt = <3300000>;
+ regulator-max-microvolt = <3300000>;
+ regulator-name = "vcc-hdmi-dsi-sensor";
+};
+
+®_dldo2 {
+ regulator-min-microvolt = <3300000>;
+ regulator-max-microvolt = <3300000>;
+ regulator-name = "vcc-mipi";
+};
+
+®_dldo3 {
+ regulator-min-microvolt = <2800000>;
+ regulator-max-microvolt = <2800000>;
+ regulator-name = "dovdd-csi";
+};
+
+®_dldo4 {
+ regulator-min-microvolt = <3300000>;
+ regulator-max-microvolt = <3300000>;
+ regulator-name = "vcc-wifi-io";
+};
+
+®_drivevbus {
+ regulator-name = "usb0-vbus";
+ status = "okay";
+};
+
+®_eldo1 {
+ regulator-always-on;
+ regulator-min-microvolt = <1800000>;
+ regulator-max-microvolt = <1800000>;
+ regulator-name = "cpvdd";
+};
+
+®_eldo3 {
+ regulator-min-microvolt = <1800000>;
+ regulator-max-microvolt = <1800000>;
+ regulator-name = "dvdd-csi";
+};
+
+®_fldo1 {
+ regulator-min-microvolt = <1200000>;
+ regulator-max-microvolt = <1200000>;
+ regulator-name = "vcc-1v2-hsic";
+};
+
+/*
+ * The A64 chip cannot work without this regulator off, although
+ * it seems to be only driving the AR100 core.
+ * Maybe we don't still know well about CPUs domain.
+ */
+®_fldo2 {
+ regulator-always-on;
+ regulator-min-microvolt = <1100000>;
+ regulator-max-microvolt = <1100000>;
+ regulator-name = "vdd-cpus";
+};
+
+®_rtc_ldo {
+ regulator-name = "vcc-rtc";
+};
+
+&uart0 {
+ pinctrl-names = "default";
+ pinctrl-0 = <&uart0_pins_a>;
+ status = "okay";
+};
+
+&usb_otg {
+ dr_mode = "otg";
+ status = "okay";
+};
+
+&usbphy {
+ usb0_id_det-gpios = <&pio 7 9 GPIO_ACTIVE_HIGH>; /* PH9 */
+ usb0_vbus-supply = <®_drivevbus>;
+ status = "okay";
+};
--
2.14.3
^ permalink raw reply related
* [PATCH 03/14] arm64: Add per-cpu infrastructure to call ARCH_WORKAROUND_2
From: Mark Rutland @ 2018-05-24 11:14 UTC (permalink / raw)
To: linux-arm-kernel
In-Reply-To: <20180522150648.28297-4-marc.zyngier@arm.com>
On Tue, May 22, 2018 at 04:06:37PM +0100, Marc Zyngier wrote:
> In a heterogeneous system, we can end up with both affected and
> unaffected CPUs. Let's check their status before calling into the
> firmware.
>
> Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Ah, I guess this may fix the issue I noted with the prior patch,
assuming we only set arm64_ssbd_callback_required for a CPU when the FW
supports the mitigation.
If so, if you fold this together with the prior patch:
Reviewed-by: Mark Rutland <mark.rutland@arm.com>
Thanks,
Mark.
> ---
> arch/arm64/kernel/cpu_errata.c | 2 ++
> arch/arm64/kernel/entry.S | 11 +++++++----
> 2 files changed, 9 insertions(+), 4 deletions(-)
>
> diff --git a/arch/arm64/kernel/cpu_errata.c b/arch/arm64/kernel/cpu_errata.c
> index 46b3aafb631a..0288d6cf560e 100644
> --- a/arch/arm64/kernel/cpu_errata.c
> +++ b/arch/arm64/kernel/cpu_errata.c
> @@ -233,6 +233,8 @@ enable_smccc_arch_workaround_1(const struct arm64_cpu_capabilities *entry)
> #endif /* CONFIG_HARDEN_BRANCH_PREDICTOR */
>
> #ifdef CONFIG_ARM64_SSBD
> +DEFINE_PER_CPU_READ_MOSTLY(u64, arm64_ssbd_callback_required);
> +
> void __init arm64_update_smccc_conduit(struct alt_instr *alt,
> __le32 *origptr, __le32 *updptr,
> int nr_inst)
> diff --git a/arch/arm64/kernel/entry.S b/arch/arm64/kernel/entry.S
> index f33e6aed3037..29ad672a6abd 100644
> --- a/arch/arm64/kernel/entry.S
> +++ b/arch/arm64/kernel/entry.S
> @@ -140,8 +140,10 @@ alternative_else_nop_endif
>
> // This macro corrupts x0-x3. It is the caller's duty
> // to save/restore them if required.
> - .macro apply_ssbd, state
> + .macro apply_ssbd, state, targ, tmp1, tmp2
> #ifdef CONFIG_ARM64_SSBD
> + ldr_this_cpu \tmp2, arm64_ssbd_callback_required, \tmp1
> + cbz \tmp2, \targ
> mov w0, #ARM_SMCCC_ARCH_WORKAROUND_2
> mov w1, #\state
> alternative_cb arm64_update_smccc_conduit
> @@ -176,12 +178,13 @@ alternative_cb_end
> ldr x19, [tsk, #TSK_TI_FLAGS] // since we can unmask debug
> disable_step_tsk x19, x20 // exceptions when scheduling.
>
> - apply_ssbd 1
> + apply_ssbd 1, 1f, x22, x23
>
> #ifdef CONFIG_ARM64_SSBD
> ldp x0, x1, [sp, #16 * 0]
> ldp x2, x3, [sp, #16 * 1]
> #endif
> +1:
>
> mov x29, xzr // fp pointed to user-space
> .else
> @@ -323,8 +326,8 @@ alternative_if ARM64_WORKAROUND_845719
> alternative_else_nop_endif
> #endif
> 3:
> - apply_ssbd 0
> -
> + apply_ssbd 0, 5f, x0, x1
> +5:
> .endif
>
> msr elr_el1, x21 // set up the return data
> --
> 2.14.2
>
>
> _______________________________________________
> linux-arm-kernel mailing list
> linux-arm-kernel at lists.infradead.org
> http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
^ permalink raw reply
* Nokia N900: refcount_t underflow, use after free
From: Pavel Machek @ 2018-05-24 11:05 UTC (permalink / raw)
To: linux-arm-kernel
In-Reply-To: <5aa1e2de-96d1-0e88-01ca-49df9ae25457@ti.com>
Hi!
> >>
> >> OK, see if the following fixes the issue for you, only build tested.
> >
> > Word-wrapped, so I applied by hand. And yes, the oops at boot is
> > gone. Thanks!
>
> Sorry about that, have to check my mail settings. Anyway will post the
> patch again, glad that it fixed your issue.
Any news here? AFAICT the bug creeped into v4.17-rcX in the
meantime...
Pavel
> regards
> Suman
>
> >
> > (Camera still does not work in -next... kills system. Oh well. Lets
> > debug that some other day.)
> >
> >> 8< ---------------------
> >> >From bac9a48fb646dc51f2030d676a0dbe3298c3b134 Mon Sep 17 00:00:00 2001
> >> From: Suman Anna <s-anna@ti.com>
> >> Date: Fri, 9 Mar 2018 16:39:59 -0600
> >> Subject: [PATCH] media: omap3isp: fix unbalanced dma_iommu_mapping
> >>
> >> The OMAP3 ISP driver manages its MMU mappings through the IOMMU-aware
> >> ARM DMA backend. The current code creates a dma_iommu_mapping and
> >> attaches this to the ISP device, but never detaches the mapping in
> >> either the probe failure paths or the driver remove path resulting
> >> in an unbalanced mapping refcount and a memory leak. Fix this properly.
> >>
> >> Reported-by: Pavel Machek <pavel@ucw.cz>
> >> Signed-off-by: Suman Anna <s-anna@ti.com>
> >
> > Tested-by: Pavel Machek <pavel@ucw.cz>
> > Pavel
> >
--
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 181 bytes
Desc: Digital signature
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20180524/38b64362/attachment-0001.sig>
^ permalink raw reply
* [PATCH 02/14] arm64: Call ARCH_WORKAROUND_2 on transitions between EL0 and EL1
From: Mark Rutland @ 2018-05-24 11:00 UTC (permalink / raw)
To: linux-arm-kernel
In-Reply-To: <20180522150648.28297-3-marc.zyngier@arm.com>
On Tue, May 22, 2018 at 04:06:36PM +0100, Marc Zyngier wrote:
> In order for the kernel to protect itself, let's call the SSBD mitigation
> implemented by the higher exception level (either hypervisor or firmware)
> on each transition between userspace and kernel.
>
> We must take the PSCI conduit into account in order to target the
> right exception level, hence the introduction of a runtime patching
> callback.
>
> Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
> ---
> arch/arm64/kernel/cpu_errata.c | 18 ++++++++++++++++++
> arch/arm64/kernel/entry.S | 22 ++++++++++++++++++++++
> include/linux/arm-smccc.h | 5 +++++
> 3 files changed, 45 insertions(+)
>
> diff --git a/arch/arm64/kernel/cpu_errata.c b/arch/arm64/kernel/cpu_errata.c
> index a900befadfe8..46b3aafb631a 100644
> --- a/arch/arm64/kernel/cpu_errata.c
> +++ b/arch/arm64/kernel/cpu_errata.c
> @@ -232,6 +232,24 @@ enable_smccc_arch_workaround_1(const struct arm64_cpu_capabilities *entry)
> }
> #endif /* CONFIG_HARDEN_BRANCH_PREDICTOR */
>
> +#ifdef CONFIG_ARM64_SSBD
> +void __init arm64_update_smccc_conduit(struct alt_instr *alt,
> + __le32 *origptr, __le32 *updptr,
> + int nr_inst)
> +{
> + u32 insn;
> +
> + BUG_ON(nr_inst != 1);
> +
> + if (psci_ops.conduit == PSCI_CONDUIT_HVC)
> + insn = aarch64_insn_get_hvc_value();
> + else
> + insn = aarch64_insn_get_smc_value();
Shouldn't this also handle the case where there is no conduit?
See below comment in apply_ssbd for rationale.
> +
> + *updptr = cpu_to_le32(insn);
> +}
> +#endif /* CONFIG_ARM64_SSBD */
> +
> #define CAP_MIDR_RANGE(model, v_min, r_min, v_max, r_max) \
> .matches = is_affected_midr_range, \
> .midr_range = MIDR_RANGE(model, v_min, r_min, v_max, r_max)
> diff --git a/arch/arm64/kernel/entry.S b/arch/arm64/kernel/entry.S
> index ec2ee720e33e..f33e6aed3037 100644
> --- a/arch/arm64/kernel/entry.S
> +++ b/arch/arm64/kernel/entry.S
> @@ -18,6 +18,7 @@
> * along with this program. If not, see <http://www.gnu.org/licenses/>.
> */
>
> +#include <linux/arm-smccc.h>
> #include <linux/init.h>
> #include <linux/linkage.h>
>
> @@ -137,6 +138,18 @@ alternative_else_nop_endif
> add \dst, \dst, #(\sym - .entry.tramp.text)
> .endm
>
> + // This macro corrupts x0-x3. It is the caller's duty
> + // to save/restore them if required.
> + .macro apply_ssbd, state
> +#ifdef CONFIG_ARM64_SSBD
> + mov w0, #ARM_SMCCC_ARCH_WORKAROUND_2
> + mov w1, #\state
> +alternative_cb arm64_update_smccc_conduit
> + nop // Patched to SMC/HVC #0
> +alternative_cb_end
> +#endif
> + .endm
If my system doesn't have SMCCC1.1, or the FW doesn't have an
implementation of ARCH_WORKAROUND_2, does this stay as a NOP?
It looks like this would be patched to an SMC, which would be fatal on
systems without EL3 FW.
> +
> .macro kernel_entry, el, regsize = 64
> .if \regsize == 32
> mov w0, w0 // zero upper 32 bits of x0
> @@ -163,6 +176,13 @@ alternative_else_nop_endif
> ldr x19, [tsk, #TSK_TI_FLAGS] // since we can unmask debug
> disable_step_tsk x19, x20 // exceptions when scheduling.
>
> + apply_ssbd 1
... and thus kernel_entry would be fatal.
Thanks,
Mark.
^ permalink raw reply
* [PATCH 9/9] arm64: bitops: Include <asm-generic/bitops/ext2-atomic-setbit.h>
From: Will Deacon @ 2018-05-24 10:59 UTC (permalink / raw)
To: linux-arm-kernel
In-Reply-To: <1527159586-8578-1-git-send-email-will.deacon@arm.com>
asm-generic/bitops/ext2-atomic-setbit.h provides the ext2 atomic bitop
definitions, so we don't need to define our own.
Signed-off-by: Will Deacon <will.deacon@arm.com>
---
arch/arm64/include/asm/bitops.h | 7 +------
1 file changed, 1 insertion(+), 6 deletions(-)
diff --git a/arch/arm64/include/asm/bitops.h b/arch/arm64/include/asm/bitops.h
index 13501460be6b..10d536b1af74 100644
--- a/arch/arm64/include/asm/bitops.h
+++ b/arch/arm64/include/asm/bitops.h
@@ -38,11 +38,6 @@
#include <asm-generic/bitops/lock.h>
#include <asm-generic/bitops/non-atomic.h>
#include <asm-generic/bitops/le.h>
-
-/*
- * Ext2 is defined to use little-endian byte ordering.
- */
-#define ext2_set_bit_atomic(lock, nr, p) test_and_set_bit_le(nr, p)
-#define ext2_clear_bit_atomic(lock, nr, p) test_and_clear_bit_le(nr, p)
+#include <asm-generic/bitops/ext2-atomic-setbit.h>
#endif /* __ASM_BITOPS_H */
--
2.1.4
^ permalink raw reply related
* [PATCH 8/9] arm64: Replace our atomic/lock bitop implementations with asm-generic
From: Will Deacon @ 2018-05-24 10:59 UTC (permalink / raw)
To: linux-arm-kernel
In-Reply-To: <1527159586-8578-1-git-send-email-will.deacon@arm.com>
The asm-generic/bitops/{atomic,lock}.h implementations are built around
the atomic-fetch ops, which we implement efficiently for both LSE and
LL/SC systems. Use that instead of our hand-rolled, out-of-line bitops.S.
Signed-off-by: Will Deacon <will.deacon@arm.com>
---
arch/arm64/include/asm/bitops.h | 14 ++------
arch/arm64/lib/Makefile | 2 +-
arch/arm64/lib/bitops.S | 76 -----------------------------------------
3 files changed, 3 insertions(+), 89 deletions(-)
delete mode 100644 arch/arm64/lib/bitops.S
diff --git a/arch/arm64/include/asm/bitops.h b/arch/arm64/include/asm/bitops.h
index 9c19594ce7cb..13501460be6b 100644
--- a/arch/arm64/include/asm/bitops.h
+++ b/arch/arm64/include/asm/bitops.h
@@ -17,22 +17,11 @@
#define __ASM_BITOPS_H
#include <linux/compiler.h>
-#include <asm/barrier.h>
#ifndef _LINUX_BITOPS_H
#error only <linux/bitops.h> can be included directly
#endif
-/*
- * Little endian assembly atomic bitops.
- */
-extern void set_bit(int nr, volatile unsigned long *p);
-extern void clear_bit(int nr, volatile unsigned long *p);
-extern void change_bit(int nr, volatile unsigned long *p);
-extern int test_and_set_bit(int nr, volatile unsigned long *p);
-extern int test_and_clear_bit(int nr, volatile unsigned long *p);
-extern int test_and_change_bit(int nr, volatile unsigned long *p);
-
#include <asm-generic/bitops/builtin-__ffs.h>
#include <asm-generic/bitops/builtin-ffs.h>
#include <asm-generic/bitops/builtin-__fls.h>
@@ -44,8 +33,9 @@ extern int test_and_change_bit(int nr, volatile unsigned long *p);
#include <asm-generic/bitops/sched.h>
#include <asm-generic/bitops/hweight.h>
-#include <asm-generic/bitops/lock.h>
+#include <asm-generic/bitops/atomic.h>
+#include <asm-generic/bitops/lock.h>
#include <asm-generic/bitops/non-atomic.h>
#include <asm-generic/bitops/le.h>
diff --git a/arch/arm64/lib/Makefile b/arch/arm64/lib/Makefile
index 137710f4dac3..68755fd70dcf 100644
--- a/arch/arm64/lib/Makefile
+++ b/arch/arm64/lib/Makefile
@@ -1,5 +1,5 @@
# SPDX-License-Identifier: GPL-2.0
-lib-y := bitops.o clear_user.o delay.o copy_from_user.o \
+lib-y := clear_user.o delay.o copy_from_user.o \
copy_to_user.o copy_in_user.o copy_page.o \
clear_page.o memchr.o memcpy.o memmove.o memset.o \
memcmp.o strcmp.o strncmp.o strlen.o strnlen.o \
diff --git a/arch/arm64/lib/bitops.S b/arch/arm64/lib/bitops.S
deleted file mode 100644
index 43ac736baa5b..000000000000
--- a/arch/arm64/lib/bitops.S
+++ /dev/null
@@ -1,76 +0,0 @@
-/*
- * Based on arch/arm/lib/bitops.h
- *
- * Copyright (C) 2013 ARM Ltd.
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License version 2 as
- * published by the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
- * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program. If not, see <http://www.gnu.org/licenses/>.
- */
-
-#include <linux/linkage.h>
-#include <asm/assembler.h>
-#include <asm/lse.h>
-
-/*
- * x0: bits 5:0 bit offset
- * bits 31:6 word offset
- * x1: address
- */
- .macro bitop, name, llsc, lse
-ENTRY( \name )
- and w3, w0, #63 // Get bit offset
- eor w0, w0, w3 // Clear low bits
- mov x2, #1
- add x1, x1, x0, lsr #3 // Get word offset
-alt_lse " prfm pstl1strm, [x1]", "nop"
- lsl x3, x2, x3 // Create mask
-
-alt_lse "1: ldxr x2, [x1]", "\lse x3, [x1]"
-alt_lse " \llsc x2, x2, x3", "nop"
-alt_lse " stxr w0, x2, [x1]", "nop"
-alt_lse " cbnz w0, 1b", "nop"
-
- ret
-ENDPROC(\name )
- .endm
-
- .macro testop, name, llsc, lse
-ENTRY( \name )
- and w3, w0, #63 // Get bit offset
- eor w0, w0, w3 // Clear low bits
- mov x2, #1
- add x1, x1, x0, lsr #3 // Get word offset
-alt_lse " prfm pstl1strm, [x1]", "nop"
- lsl x4, x2, x3 // Create mask
-
-alt_lse "1: ldxr x2, [x1]", "\lse x4, x2, [x1]"
- lsr x0, x2, x3
-alt_lse " \llsc x2, x2, x4", "nop"
-alt_lse " stlxr w5, x2, [x1]", "nop"
-alt_lse " cbnz w5, 1b", "nop"
-alt_lse " dmb ish", "nop"
-
- and x0, x0, #1
- ret
-ENDPROC(\name )
- .endm
-
-/*
- * Atomic bit operations.
- */
- bitop change_bit, eor, steor
- bitop clear_bit, bic, stclr
- bitop set_bit, orr, stset
-
- testop test_and_change_bit, eor, ldeoral
- testop test_and_clear_bit, bic, ldclral
- testop test_and_set_bit, orr, ldsetal
--
2.1.4
^ permalink raw reply related
* [PATCH 7/9] asm-generic/bitops/lock.h: Rewrite using atomic_fetch_*
From: Will Deacon @ 2018-05-24 10:59 UTC (permalink / raw)
To: linux-arm-kernel
In-Reply-To: <1527159586-8578-1-git-send-email-will.deacon@arm.com>
The lock bitops can be implemented more efficiently using the atomic_fetch_*
ops, which provide finer-grained control over the memory ordering semantics
than the bitops.
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Ingo Molnar <mingo@kernel.org>
Signed-off-by: Will Deacon <will.deacon@arm.com>
---
include/asm-generic/bitops/lock.h | 68 ++++++++++++++++++++++++++++++++-------
1 file changed, 56 insertions(+), 12 deletions(-)
diff --git a/include/asm-generic/bitops/lock.h b/include/asm-generic/bitops/lock.h
index 67ab280ad134..3ae021368f48 100644
--- a/include/asm-generic/bitops/lock.h
+++ b/include/asm-generic/bitops/lock.h
@@ -2,6 +2,10 @@
#ifndef _ASM_GENERIC_BITOPS_LOCK_H_
#define _ASM_GENERIC_BITOPS_LOCK_H_
+#include <linux/atomic.h>
+#include <linux/compiler.h>
+#include <asm/barrier.h>
+
/**
* test_and_set_bit_lock - Set a bit and return its old value, for lock
* @nr: Bit to set
@@ -11,7 +15,20 @@
* the returned value is 0.
* It can be used to implement bit locks.
*/
-#define test_and_set_bit_lock(nr, addr) test_and_set_bit(nr, addr)
+static inline int test_and_set_bit_lock(unsigned int nr,
+ volatile unsigned long *p)
+{
+ long old;
+ unsigned long mask = BIT_MASK(nr);
+
+ p += BIT_WORD(nr);
+ if (READ_ONCE(*p) & mask)
+ return 1;
+
+ old = atomic_long_fetch_or_acquire(mask, (atomic_long_t *)p);
+ return !!(old & mask);
+}
+
/**
* clear_bit_unlock - Clear a bit in memory, for unlock
@@ -20,11 +37,11 @@
*
* This operation is atomic and provides release barrier semantics.
*/
-#define clear_bit_unlock(nr, addr) \
-do { \
- smp_mb__before_atomic(); \
- clear_bit(nr, addr); \
-} while (0)
+static inline void clear_bit_unlock(unsigned int nr, volatile unsigned long *p)
+{
+ p += BIT_WORD(nr);
+ atomic_long_fetch_andnot_release(BIT_MASK(nr), (atomic_long_t *)p);
+}
/**
* __clear_bit_unlock - Clear a bit in memory, for unlock
@@ -37,11 +54,38 @@ do { \
*
* See for example x86's implementation.
*/
-#define __clear_bit_unlock(nr, addr) \
-do { \
- smp_mb__before_atomic(); \
- clear_bit(nr, addr); \
-} while (0)
+static inline void __clear_bit_unlock(unsigned int nr,
+ volatile unsigned long *p)
+{
+ unsigned long old;
-#endif /* _ASM_GENERIC_BITOPS_LOCK_H_ */
+ p += BIT_WORD(nr);
+ old = READ_ONCE(*p);
+ old &= ~BIT_MASK(nr);
+ atomic_long_set_release((atomic_long_t *)p, old);
+}
+
+/**
+ * clear_bit_unlock_is_negative_byte - Clear a bit in memory and test if bottom
+ * byte is negative, for unlock.
+ * @nr: the bit to clear
+ * @addr: the address to start counting from
+ *
+ * This is a bit of a one-trick-pony for the filemap code, which clears
+ * PG_locked and tests PG_waiters,
+ */
+#ifndef clear_bit_unlock_is_negative_byte
+static inline bool clear_bit_unlock_is_negative_byte(unsigned int nr,
+ volatile unsigned long *p)
+{
+ long old;
+ unsigned long mask = BIT_MASK(nr);
+
+ p += BIT_WORD(nr);
+ old = atomic_long_fetch_andnot_release(mask, (atomic_long_t *)p);
+ return !!(old & BIT(7));
+}
+#define clear_bit_unlock_is_negative_byte clear_bit_unlock_is_negative_byte
+#endif
+#endif /* _ASM_GENERIC_BITOPS_LOCK_H_ */
--
2.1.4
^ permalink raw reply related
* [PATCH 6/9] asm-generic/bitops/atomic.h: Rewrite using atomic_fetch_*
From: Will Deacon @ 2018-05-24 10:59 UTC (permalink / raw)
To: linux-arm-kernel
In-Reply-To: <1527159586-8578-1-git-send-email-will.deacon@arm.com>
The atomic bitops can actually be implemented pretty efficiently using
the atomic_fetch_* ops, rather than explicit use of spinlocks.
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Ingo Molnar <mingo@kernel.org>
Signed-off-by: Will Deacon <will.deacon@arm.com>
---
include/asm-generic/bitops/atomic.h | 188 +++++++-----------------------------
1 file changed, 33 insertions(+), 155 deletions(-)
diff --git a/include/asm-generic/bitops/atomic.h b/include/asm-generic/bitops/atomic.h
index 04deffaf5f7d..bca92586c2f6 100644
--- a/include/asm-generic/bitops/atomic.h
+++ b/include/asm-generic/bitops/atomic.h
@@ -2,189 +2,67 @@
#ifndef _ASM_GENERIC_BITOPS_ATOMIC_H_
#define _ASM_GENERIC_BITOPS_ATOMIC_H_
-#include <asm/types.h>
-#include <linux/irqflags.h>
-
-#ifdef CONFIG_SMP
-#include <asm/spinlock.h>
-#include <asm/cache.h> /* we use L1_CACHE_BYTES */
-
-/* Use an array of spinlocks for our atomic_ts.
- * Hash function to index into a different SPINLOCK.
- * Since "a" is usually an address, use one spinlock per cacheline.
- */
-# define ATOMIC_HASH_SIZE 4
-# define ATOMIC_HASH(a) (&(__atomic_hash[ (((unsigned long) a)/L1_CACHE_BYTES) & (ATOMIC_HASH_SIZE-1) ]))
-
-extern arch_spinlock_t __atomic_hash[ATOMIC_HASH_SIZE] __lock_aligned;
-
-/* Can't use raw_spin_lock_irq because of #include problems, so
- * this is the substitute */
-#define _atomic_spin_lock_irqsave(l,f) do { \
- arch_spinlock_t *s = ATOMIC_HASH(l); \
- local_irq_save(f); \
- arch_spin_lock(s); \
-} while(0)
-
-#define _atomic_spin_unlock_irqrestore(l,f) do { \
- arch_spinlock_t *s = ATOMIC_HASH(l); \
- arch_spin_unlock(s); \
- local_irq_restore(f); \
-} while(0)
-
-
-#else
-# define _atomic_spin_lock_irqsave(l,f) do { local_irq_save(f); } while (0)
-# define _atomic_spin_unlock_irqrestore(l,f) do { local_irq_restore(f); } while (0)
-#endif
+#include <linux/atomic.h>
+#include <linux/compiler.h>
+#include <asm/barrier.h>
/*
- * NMI events can occur at any time, including when interrupts have been
- * disabled by *_irqsave(). So you can get NMI events occurring while a
- * *_bit function is holding a spin lock. If the NMI handler also wants
- * to do bit manipulation (and they do) then you can get a deadlock
- * between the original caller of *_bit() and the NMI handler.
- *
- * by Keith Owens
+ * Implementation of atomic bitops using atomic-fetch ops.
+ * See Documentation/atomic_bitops.txt for details.
*/
-/**
- * set_bit - Atomically set a bit in memory
- * @nr: the bit to set
- * @addr: the address to start counting from
- *
- * This function is atomic and may not be reordered. See __set_bit()
- * if you do not require the atomic guarantees.
- *
- * Note: there are no guarantees that this function will not be reordered
- * on non x86 architectures, so if you are writing portable code,
- * make sure not to rely on its reordering guarantees.
- *
- * Note that @nr may be almost arbitrarily large; this function is not
- * restricted to acting on a single-word quantity.
- */
-static inline void set_bit(int nr, volatile unsigned long *addr)
+static inline void set_bit(unsigned int nr, volatile unsigned long *p)
{
- unsigned long mask = BIT_MASK(nr);
- unsigned long *p = ((unsigned long *)addr) + BIT_WORD(nr);
- unsigned long flags;
-
- _atomic_spin_lock_irqsave(p, flags);
- *p |= mask;
- _atomic_spin_unlock_irqrestore(p, flags);
+ p += BIT_WORD(nr);
+ atomic_long_fetch_or_relaxed(BIT_MASK(nr), (atomic_long_t *)p);
}
-/**
- * clear_bit - Clears a bit in memory
- * @nr: Bit to clear
- * @addr: Address to start counting from
- *
- * clear_bit() is atomic and may not be reordered. However, it does
- * not contain a memory barrier, so if it is used for locking purposes,
- * you should call smp_mb__before_atomic() and/or smp_mb__after_atomic()
- * in order to ensure changes are visible on other processors.
- */
-static inline void clear_bit(int nr, volatile unsigned long *addr)
+static inline void clear_bit(unsigned int nr, volatile unsigned long *p)
{
- unsigned long mask = BIT_MASK(nr);
- unsigned long *p = ((unsigned long *)addr) + BIT_WORD(nr);
- unsigned long flags;
-
- _atomic_spin_lock_irqsave(p, flags);
- *p &= ~mask;
- _atomic_spin_unlock_irqrestore(p, flags);
+ p += BIT_WORD(nr);
+ atomic_long_fetch_andnot_relaxed(BIT_MASK(nr), (atomic_long_t *)p);
}
-/**
- * change_bit - Toggle a bit in memory
- * @nr: Bit to change
- * @addr: Address to start counting from
- *
- * change_bit() is atomic and may not be reordered. It may be
- * reordered on other architectures than x86.
- * Note that @nr may be almost arbitrarily large; this function is not
- * restricted to acting on a single-word quantity.
- */
-static inline void change_bit(int nr, volatile unsigned long *addr)
+static inline void change_bit(unsigned int nr, volatile unsigned long *p)
{
- unsigned long mask = BIT_MASK(nr);
- unsigned long *p = ((unsigned long *)addr) + BIT_WORD(nr);
- unsigned long flags;
-
- _atomic_spin_lock_irqsave(p, flags);
- *p ^= mask;
- _atomic_spin_unlock_irqrestore(p, flags);
+ p += BIT_WORD(nr);
+ atomic_long_fetch_xor_relaxed(BIT_MASK(nr), (atomic_long_t *)p);
}
-/**
- * test_and_set_bit - Set a bit and return its old value
- * @nr: Bit to set
- * @addr: Address to count from
- *
- * This operation is atomic and cannot be reordered.
- * It may be reordered on other architectures than x86.
- * It also implies a memory barrier.
- */
-static inline int test_and_set_bit(int nr, volatile unsigned long *addr)
+static inline int test_and_set_bit(unsigned int nr, volatile unsigned long *p)
{
+ long old;
unsigned long mask = BIT_MASK(nr);
- unsigned long *p = ((unsigned long *)addr) + BIT_WORD(nr);
- unsigned long old;
- unsigned long flags;
- _atomic_spin_lock_irqsave(p, flags);
- old = *p;
- *p = old | mask;
- _atomic_spin_unlock_irqrestore(p, flags);
+ p += BIT_WORD(nr);
+ if (READ_ONCE(*p) & mask)
+ return 1;
- return (old & mask) != 0;
+ old = atomic_long_fetch_or(mask, (atomic_long_t *)p);
+ return !!(old & mask);
}
-/**
- * test_and_clear_bit - Clear a bit and return its old value
- * @nr: Bit to clear
- * @addr: Address to count from
- *
- * This operation is atomic and cannot be reordered.
- * It can be reorderdered on other architectures other than x86.
- * It also implies a memory barrier.
- */
-static inline int test_and_clear_bit(int nr, volatile unsigned long *addr)
+static inline int test_and_clear_bit(unsigned int nr, volatile unsigned long *p)
{
+ long old;
unsigned long mask = BIT_MASK(nr);
- unsigned long *p = ((unsigned long *)addr) + BIT_WORD(nr);
- unsigned long old;
- unsigned long flags;
- _atomic_spin_lock_irqsave(p, flags);
- old = *p;
- *p = old & ~mask;
- _atomic_spin_unlock_irqrestore(p, flags);
+ p += BIT_WORD(nr);
+ if (!(READ_ONCE(*p) & mask))
+ return 0;
- return (old & mask) != 0;
+ old = atomic_long_fetch_andnot(mask, (atomic_long_t *)p);
+ return !!(old & mask);
}
-/**
- * test_and_change_bit - Change a bit and return its old value
- * @nr: Bit to change
- * @addr: Address to count from
- *
- * This operation is atomic and cannot be reordered.
- * It also implies a memory barrier.
- */
-static inline int test_and_change_bit(int nr, volatile unsigned long *addr)
+static inline int test_and_change_bit(unsigned int nr, volatile unsigned long *p)
{
+ long old;
unsigned long mask = BIT_MASK(nr);
- unsigned long *p = ((unsigned long *)addr) + BIT_WORD(nr);
- unsigned long old;
- unsigned long flags;
-
- _atomic_spin_lock_irqsave(p, flags);
- old = *p;
- *p = old ^ mask;
- _atomic_spin_unlock_irqrestore(p, flags);
- return (old & mask) != 0;
+ p += BIT_WORD(nr);
+ old = atomic_long_fetch_xor(mask, (atomic_long_t *)p);
+ return !!(old & mask);
}
#endif /* _ASM_GENERIC_BITOPS_ATOMIC_H */
--
2.1.4
^ permalink raw reply related
* [PATCH 5/9] sh: Don't pull in all of linux/bitops.h in asm/cmpxchg-xchg.h
From: Will Deacon @ 2018-05-24 10:59 UTC (permalink / raw)
To: linux-arm-kernel
In-Reply-To: <1527159586-8578-1-git-send-email-will.deacon@arm.com>
The sh implementation of asm/cmpxchg-xchg.h pulls in linux/bitops.h
so that it can refer to BITS_PER_BYTE. It also transitively relies on
this pulling in linux/compiler.h for READ_ONCE.
Replace the #include with linux/bits.h and linux/compiler.h
Signed-off-by: Will Deacon <will.deacon@arm.com>
---
arch/sh/include/asm/cmpxchg-xchg.h | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/arch/sh/include/asm/cmpxchg-xchg.h b/arch/sh/include/asm/cmpxchg-xchg.h
index 1e881f5db659..593a9704782b 100644
--- a/arch/sh/include/asm/cmpxchg-xchg.h
+++ b/arch/sh/include/asm/cmpxchg-xchg.h
@@ -8,7 +8,8 @@
* This work is licensed under the terms of the GNU GPL, version 2. See the
* file "COPYING" in the main directory of this archive for more details.
*/
-#include <linux/bitops.h>
+#include <linux/bits.h>
+#include <linux/compiler.h>
#include <asm/byteorder.h>
/*
--
2.1.4
^ permalink raw reply related
* [PATCH 4/9] openrisc: Don't pull in all of linux/bitops.h in asm/cmpxchg.h
From: Will Deacon @ 2018-05-24 10:59 UTC (permalink / raw)
To: linux-arm-kernel
In-Reply-To: <1527159586-8578-1-git-send-email-will.deacon@arm.com>
The openrisc implementation of asm/cmpxchg.h pulls in linux/bitops.h
so that it can refer to BITS_PER_BYTE. It also transitively relies on
this pulling in linux/compiler.h for READ_ONCE.
Replace the #include with linux/bits.h and linux/compiler.h
Signed-off-by: Will Deacon <will.deacon@arm.com>
---
arch/openrisc/include/asm/cmpxchg.h | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/arch/openrisc/include/asm/cmpxchg.h b/arch/openrisc/include/asm/cmpxchg.h
index d29f7db53906..f9cd43a39d72 100644
--- a/arch/openrisc/include/asm/cmpxchg.h
+++ b/arch/openrisc/include/asm/cmpxchg.h
@@ -16,8 +16,9 @@
#ifndef __ASM_OPENRISC_CMPXCHG_H
#define __ASM_OPENRISC_CMPXCHG_H
+#include <linux/bits.h>
+#include <linux/compiler.h>
#include <linux/types.h>
-#include <linux/bitops.h>
#define __HAVE_ARCH_CMPXCHG 1
--
2.1.4
^ permalink raw reply related
* [PATCH 3/9] asm-generic: Move some macros from linux/bitops.h to a new bits.h file
From: Will Deacon @ 2018-05-24 10:59 UTC (permalink / raw)
To: linux-arm-kernel
In-Reply-To: <1527159586-8578-1-git-send-email-will.deacon@arm.com>
In preparation for implementing the asm-generic atomic bitops in terms
of atomic_long_*, we need to prevent asm/atomic.h implementations from
pulling in linux/bitops.h. A common reason for this include is for the
BITS_PER_BYTE definition, so move this and some other BIT and masking
macros into a new header file, linux/bits.h
Signed-off-by: Will Deacon <will.deacon@arm.com>
---
include/linux/bitops.h | 22 +---------------------
include/linux/bits.h | 26 ++++++++++++++++++++++++++
2 files changed, 27 insertions(+), 21 deletions(-)
create mode 100644 include/linux/bits.h
diff --git a/include/linux/bitops.h b/include/linux/bitops.h
index 4cac4e1a72ff..af419012d77d 100644
--- a/include/linux/bitops.h
+++ b/include/linux/bitops.h
@@ -2,29 +2,9 @@
#ifndef _LINUX_BITOPS_H
#define _LINUX_BITOPS_H
#include <asm/types.h>
+#include <linux/bits.h>
-#ifdef __KERNEL__
-#define BIT(nr) (1UL << (nr))
-#define BIT_ULL(nr) (1ULL << (nr))
-#define BIT_MASK(nr) (1UL << ((nr) % BITS_PER_LONG))
-#define BIT_WORD(nr) ((nr) / BITS_PER_LONG)
-#define BIT_ULL_MASK(nr) (1ULL << ((nr) % BITS_PER_LONG_LONG))
-#define BIT_ULL_WORD(nr) ((nr) / BITS_PER_LONG_LONG)
-#define BITS_PER_BYTE 8
#define BITS_TO_LONGS(nr) DIV_ROUND_UP(nr, BITS_PER_BYTE * sizeof(long))
-#endif
-
-/*
- * Create a contiguous bitmask starting@bit position @l and ending at
- * position @h. For example
- * GENMASK_ULL(39, 21) gives us the 64bit vector 0x000000ffffe00000.
- */
-#define GENMASK(h, l) \
- (((~0UL) - (1UL << (l)) + 1) & (~0UL >> (BITS_PER_LONG - 1 - (h))))
-
-#define GENMASK_ULL(h, l) \
- (((~0ULL) - (1ULL << (l)) + 1) & \
- (~0ULL >> (BITS_PER_LONG_LONG - 1 - (h))))
extern unsigned int __sw_hweight8(unsigned int w);
extern unsigned int __sw_hweight16(unsigned int w);
diff --git a/include/linux/bits.h b/include/linux/bits.h
new file mode 100644
index 000000000000..2b7b532c1d51
--- /dev/null
+++ b/include/linux/bits.h
@@ -0,0 +1,26 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef __LINUX_BITS_H
+#define __LINUX_BITS_H
+#include <asm/bitsperlong.h>
+
+#define BIT(nr) (1UL << (nr))
+#define BIT_ULL(nr) (1ULL << (nr))
+#define BIT_MASK(nr) (1UL << ((nr) % BITS_PER_LONG))
+#define BIT_WORD(nr) ((nr) / BITS_PER_LONG)
+#define BIT_ULL_MASK(nr) (1ULL << ((nr) % BITS_PER_LONG_LONG))
+#define BIT_ULL_WORD(nr) ((nr) / BITS_PER_LONG_LONG)
+#define BITS_PER_BYTE 8
+
+/*
+ * Create a contiguous bitmask starting@bit position @l and ending at
+ * position @h. For example
+ * GENMASK_ULL(39, 21) gives us the 64bit vector 0x000000ffffe00000.
+ */
+#define GENMASK(h, l) \
+ (((~0UL) - (1UL << (l)) + 1) & (~0UL >> (BITS_PER_LONG - 1 - (h))))
+
+#define GENMASK_ULL(h, l) \
+ (((~0ULL) - (1ULL << (l)) + 1) & \
+ (~0ULL >> (BITS_PER_LONG_LONG - 1 - (h))))
+
+#endif /* __LINUX_BITS_H */
--
2.1.4
^ permalink raw reply related
* [PATCH 2/9] m68k: Don't use asm-generic/bitops/lock.h
From: Will Deacon @ 2018-05-24 10:59 UTC (permalink / raw)
To: linux-arm-kernel
In-Reply-To: <1527159586-8578-1-git-send-email-will.deacon@arm.com>
asm-generic/bitops/lock.h is shortly going to be built on top of the
atomic_long_* API, which introduces a nasty circular dependency for
m68k where linux/atomic.h pulls in linux/bitops.h via:
linux/atomic.h
asm/atomic.h
linux/irqflags.h
asm/irqflags.h
linux/preempt.h
asm/preempt.h
asm-generic/preempt.h
linux/thread_info.h
asm/thread_info.h
asm/page.h
asm-generic/getorder.h
linux/log2.h
linux/bitops.h
Since m68k isn't SMP and doesn't support ACQUIRE/RELEASE barriers, we
can just define the lock bitops in terms of the atomic bitops in the
asm/bitops.h header.
Acked-by: Geert Uytterhoeven <geert@linux-m68k.org>
Signed-off-by: Will Deacon <will.deacon@arm.com>
---
arch/m68k/include/asm/bitops.h | 6 +++++-
1 file changed, 5 insertions(+), 1 deletion(-)
diff --git a/arch/m68k/include/asm/bitops.h b/arch/m68k/include/asm/bitops.h
index 93b47b1f6fb4..18193419f97d 100644
--- a/arch/m68k/include/asm/bitops.h
+++ b/arch/m68k/include/asm/bitops.h
@@ -515,12 +515,16 @@ static inline int __fls(int x)
#endif
+/* Simple test-and-set bit locks */
+#define test_and_set_bit_lock test_and_set_bit
+#define clear_bit_unlock clear_bit
+#define __clear_bit_unlock clear_bit_unlock
+
#include <asm-generic/bitops/ext2-atomic.h>
#include <asm-generic/bitops/le.h>
#include <asm-generic/bitops/fls64.h>
#include <asm-generic/bitops/sched.h>
#include <asm-generic/bitops/hweight.h>
-#include <asm-generic/bitops/lock.h>
#endif /* __KERNEL__ */
#endif /* _M68K_BITOPS_H */
--
2.1.4
^ permalink raw reply related
* [PATCH 1/9] h8300: Don't include linux/kernel.h in asm/atomic.h
From: Will Deacon @ 2018-05-24 10:59 UTC (permalink / raw)
To: linux-arm-kernel
In-Reply-To: <1527159586-8578-1-git-send-email-will.deacon@arm.com>
linux/kernel.h isn't needed by asm/atomic.h and will result in circular
dependencies when the asm-generic atomic bitops are built around the
tomic_long_t interface.
Remove the broad include and replace it with linux/compiler.h for
READ_ONCE etc and asm/irqflags.h for arch_local_irq_save etc.
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Signed-off-by: Will Deacon <will.deacon@arm.com>
---
arch/h8300/include/asm/atomic.h | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/arch/h8300/include/asm/atomic.h b/arch/h8300/include/asm/atomic.h
index 941e7554e886..b174dec099bf 100644
--- a/arch/h8300/include/asm/atomic.h
+++ b/arch/h8300/include/asm/atomic.h
@@ -2,8 +2,10 @@
#ifndef __ARCH_H8300_ATOMIC__
#define __ARCH_H8300_ATOMIC__
+#include <linux/compiler.h>
#include <linux/types.h>
#include <asm/cmpxchg.h>
+#include <asm/irqflags.h>
/*
* Atomic operations that C can't guarantee us. Useful for
@@ -15,8 +17,6 @@
#define atomic_read(v) READ_ONCE((v)->counter)
#define atomic_set(v, i) WRITE_ONCE(((v)->counter), (i))
-#include <linux/kernel.h>
-
#define ATOMIC_OP_RETURN(op, c_op) \
static inline int atomic_##op##_return(int i, atomic_t *v) \
{ \
--
2.1.4
^ permalink raw reply related
* [PATCH 0/9] Rewrite asm-generic/bitops/{atomic, lock}.h and use on arm64
From: Will Deacon @ 2018-05-24 10:59 UTC (permalink / raw)
To: linux-arm-kernel
Hi all,
This patch series has previously been posted in RFC form here:
RFCv1: https://www.spinics.net/lists/arm-kernel/msg634719.html
RFCv2: https://www.spinics.net/lists/arm-kernel/msg636875.html
Changes since RFCv2 include:
* Rebased onto v4.17-rc4, which allowed me to drop some patches from
the series which were merged in 4.16.
* Moved bit.h to be linux/bit.h instead of asm-generic/bit.h
Thanks,
Will
--->8
Will Deacon (9):
h8300: Don't include linux/kernel.h in asm/atomic.h
m68k: Don't use asm-generic/bitops/lock.h
asm-generic: Move some macros from linux/bitops.h to a new bits.h file
openrisc: Don't pull in all of linux/bitops.h in asm/cmpxchg.h
sh: Don't pull in all of linux/bitops.h in asm/cmpxchg-xchg.h
asm-generic/bitops/atomic.h: Rewrite using atomic_fetch_*
asm-generic/bitops/lock.h: Rewrite using atomic_fetch_*
arm64: Replace our atomic/lock bitop implementations with asm-generic
arm64: bitops: Include <asm-generic/bitops/ext2-atomic-setbit.h>
arch/arm64/include/asm/bitops.h | 21 +---
arch/arm64/lib/Makefile | 2 +-
arch/arm64/lib/bitops.S | 76 ---------------
arch/h8300/include/asm/atomic.h | 4 +-
arch/m68k/include/asm/bitops.h | 6 +-
arch/openrisc/include/asm/cmpxchg.h | 3 +-
arch/sh/include/asm/cmpxchg-xchg.h | 3 +-
include/asm-generic/bitops/atomic.h | 188 +++++++-----------------------------
include/asm-generic/bitops/lock.h | 68 ++++++++++---
include/linux/bitops.h | 22 +----
include/linux/bits.h | 26 +++++
11 files changed, 131 insertions(+), 288 deletions(-)
delete mode 100644 arch/arm64/lib/bitops.S
create mode 100644 include/linux/bits.h
--
2.1.4
^ permalink raw reply
* [PATCH 01/14] arm/arm64: smccc: Add SMCCC-specific return codes
From: Mark Rutland @ 2018-05-24 10:55 UTC (permalink / raw)
To: linux-arm-kernel
In-Reply-To: <20180522150648.28297-2-marc.zyngier@arm.com>
On Tue, May 22, 2018 at 04:06:35PM +0100, Marc Zyngier wrote:
> We've so far used the PSCI return codes for SMCCC because they
> were extremely similar. But with the new ARM DEN 0070A specification,
> "NOT_REQUIRED" (-2) is clashing with PSCI's "PSCI_RET_INVALID_PARAMS".
>
> Let's bite the bullet and add SMCCC specific return codes. Users
> can be repainted as and when required.
>
> Acked-by: Will Deacon <will.deacon@arm.com>
> Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Reviewed-by: Mark Rutland <mark.rutland@arm.com>
Mark.
> ---
> include/linux/arm-smccc.h | 5 +++++
> 1 file changed, 5 insertions(+)
>
> diff --git a/include/linux/arm-smccc.h b/include/linux/arm-smccc.h
> index a031897fca76..c89da86de99f 100644
> --- a/include/linux/arm-smccc.h
> +++ b/include/linux/arm-smccc.h
> @@ -291,5 +291,10 @@ asmlinkage void __arm_smccc_hvc(unsigned long a0, unsigned long a1,
> */
> #define arm_smccc_1_1_hvc(...) __arm_smccc_1_1(SMCCC_HVC_INST, __VA_ARGS__)
>
> +/* Return codes defined in ARM DEN 0070A */
> +#define SMCCC_RET_SUCCESS 0
> +#define SMCCC_RET_NOT_SUPPORTED -1
> +#define SMCCC_RET_NOT_REQUIRED -2
> +
> #endif /*__ASSEMBLY__*/
> #endif /*__LINUX_ARM_SMCCC_H*/
> --
> 2.14.2
>
> _______________________________________________
> kvmarm mailing list
> kvmarm at lists.cs.columbia.edu
> https://lists.cs.columbia.edu/mailman/listinfo/kvmarm
^ permalink raw reply
* [PATCH 02/14] arm64: Call ARCH_WORKAROUND_2 on transitions between EL0 and EL1
From: Mark Rutland @ 2018-05-24 10:52 UTC (permalink / raw)
To: linux-arm-kernel
In-Reply-To: <0b03ce12-510e-464b-63de-cd481d7cb1c5@arm.com>
On Wed, May 23, 2018 at 10:23:20AM +0100, Julien Grall wrote:
> Hi Marc,
>
> On 05/22/2018 04:06 PM, Marc Zyngier wrote:
> > diff --git a/arch/arm64/kernel/entry.S b/arch/arm64/kernel/entry.S
> > index ec2ee720e33e..f33e6aed3037 100644
> > --- a/arch/arm64/kernel/entry.S
> > +++ b/arch/arm64/kernel/entry.S
> > @@ -18,6 +18,7 @@
> > * along with this program. If not, see <http://www.gnu.org/licenses/>.
> > */
> > +#include <linux/arm-smccc.h>
> > #include <linux/init.h>
> > #include <linux/linkage.h>
> > @@ -137,6 +138,18 @@ alternative_else_nop_endif
> > add \dst, \dst, #(\sym - .entry.tramp.text)
> > .endm
> > + // This macro corrupts x0-x3. It is the caller's duty
> > + // to save/restore them if required.
>
> NIT: Shouldn't you use /* ... */ for multi-line comments?
There's no requirement to do so, and IIRC even Torvalds prefers '//'
comments for multi-line things these days.
Mark.
^ permalink raw reply
* [PATCH v4 1/2] arm64/sve: Thin out initialisation sanity-checks for sve_max_vl
From: Will Deacon @ 2018-05-24 10:40 UTC (permalink / raw)
To: linux-arm-kernel
In-Reply-To: <1527097616-25214-2-git-send-email-Dave.Martin@arm.com>
On Wed, May 23, 2018 at 06:46:55PM +0100, Dave Martin wrote:
> Now that the kernel SVE support is reasonably mature, it is
> excessive to default sve_max_vl to the invalid value -1 and then
> sprinkle WARN_ON()s around the place to make sure it has been
> initialised before use. The cpufeatures code already runs pretty
> early, and will ensure sve_max_vl gets initialised.
>
> This patch initialises sve_max_vl to something sane that will be
> supported by every SVE implementation, and removes most of the
> sanity checks.
>
> The checks in find_supported_vector_length() are retained for now.
> If anything goes horribly wrong, we are likely to trip a check here
> sooner or later.
>
> Signed-off-by: Dave Martin <Dave.Martin@arm.com>
> ---
> arch/arm64/kernel/fpsimd.c | 17 ++++-------------
> arch/arm64/kernel/ptrace.c | 3 ---
> 2 files changed, 4 insertions(+), 16 deletions(-)
Thanks, this makes the code a bit more readable imo:
Acked-by: Will Deacon <will.deacon@arm.com>
Will
^ permalink raw reply
* [PATCH v3 3/3] ARM: dts: Renesas R9A06G032 SMP enable method
From: Michel Pollet @ 2018-05-24 10:39 UTC (permalink / raw)
To: linux-arm-kernel
Add a special enable method for the second CA7 of the R9A06G032
as well as the default value for the "cpu-release-addr" property.
Signed-off-by: Michel Pollet <michel.pollet@bp.renesas.com>
---
arch/arm/boot/dts/r9a06g032.dtsi | 2 ++
1 file changed, 2 insertions(+)
diff --git a/arch/arm/boot/dts/r9a06g032.dtsi b/arch/arm/boot/dts/r9a06g032.dtsi
index 9534f1b..d7b5414 100644
--- a/arch/arm/boot/dts/r9a06g032.dtsi
+++ b/arch/arm/boot/dts/r9a06g032.dtsi
@@ -30,6 +30,8 @@
compatible = "arm,cortex-a7";
reg = <1>;
clocks = <&sysctrl R9A06G032_DIV_CA7>;
+ enable-method = "renesas,r9a06g032-smp";
+ cpu-release-addr = <0x4000c204>;
};
};
--
2.7.4
^ permalink raw reply related
* [PATCH v3 2/3] arm: shmobile: Add the R9A06G032 SMP enabler driver
From: Michel Pollet @ 2018-05-24 10:30 UTC (permalink / raw)
To: linux-arm-kernel
In-Reply-To: <1527157834-7747-1-git-send-email-michel.pollet@bp.renesas.com>
The Renesas R9A06G032 second CA7 is parked in a ROM pen at boot time, it
requires a special enable method to get it started.
Signed-off-by: Michel Pollet <michel.pollet@bp.renesas.com>
---
arch/arm/mach-shmobile/Makefile | 1 +
arch/arm/mach-shmobile/smp-r9a06g032.c | 85 ++++++++++++++++++++++++++++++++++
2 files changed, 86 insertions(+)
create mode 100644 arch/arm/mach-shmobile/smp-r9a06g032.c
diff --git a/arch/arm/mach-shmobile/Makefile b/arch/arm/mach-shmobile/Makefile
index 1939f52..d7fc98f 100644
--- a/arch/arm/mach-shmobile/Makefile
+++ b/arch/arm/mach-shmobile/Makefile
@@ -34,6 +34,7 @@ smp-$(CONFIG_ARCH_SH73A0) += smp-sh73a0.o headsmp-scu.o platsmp-scu.o
smp-$(CONFIG_ARCH_R8A7779) += smp-r8a7779.o headsmp-scu.o platsmp-scu.o
smp-$(CONFIG_ARCH_R8A7790) += smp-r8a7790.o
smp-$(CONFIG_ARCH_R8A7791) += smp-r8a7791.o
+smp-$(CONFIG_ARCH_R9A06G032) += smp-r9a06g032.o
smp-$(CONFIG_ARCH_EMEV2) += smp-emev2.o headsmp-scu.o platsmp-scu.o
# PM objects
diff --git a/arch/arm/mach-shmobile/smp-r9a06g032.c b/arch/arm/mach-shmobile/smp-r9a06g032.c
new file mode 100644
index 0000000..a536e89
--- /dev/null
+++ b/arch/arm/mach-shmobile/smp-r9a06g032.c
@@ -0,0 +1,85 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * RZ/N1D Second CA7 enabler.
+ *
+ * Copyright (C) 2018 Renesas Electronics Europe Limited
+ *
+ * Michel Pollet <michel.pollet@bp.renesas.com>, <buserror@gmail.com>
+ * Derived from action,s500-smp
+ */
+
+#include <linux/delay.h>
+#include <linux/io.h>
+#include <linux/of.h>
+#include <linux/of_address.h>
+#include <linux/smp.h>
+#include <asm/cacheflush.h>
+#include <asm/smp_plat.h>
+#include <asm/smp_scu.h>
+
+/*
+ * The second CPU is parked in ROM at boot time. It requires waking it after
+ * writing an address into the BOOTADDR register of sysctrl.
+ *
+ * So the default value of the "cpu-release-addr" corresponds to BOOTADDR...
+ *
+ * *However* the BOOTADDR register is not available when the kernel
+ * starts in NONSEC mode.
+ *
+ * So for NONSEC mode, the bootloader re-parks the second CPU into a pen
+ * in SRAM, and changes the "cpu-release-addr" of linux's DT to a SRAM address,
+ * which is not restricted.
+ */
+
+static void __iomem *cpu_bootaddr;
+
+static DEFINE_SPINLOCK(cpu_lock);
+
+static int rzn1_smp_boot_secondary(unsigned int cpu, struct task_struct *idle)
+{
+ if (!cpu_bootaddr)
+ return -ENODEV;
+
+ spin_lock(&cpu_lock);
+
+ writel(__pa_symbol(secondary_startup), cpu_bootaddr);
+ arch_send_wakeup_ipi_mask(cpumask_of(cpu));
+
+ spin_unlock(&cpu_lock);
+
+ return 0;
+}
+
+static void __init rzn1_smp_prepare_cpus(unsigned int max_cpus)
+{
+ struct device_node *dn;
+ int ret;
+ u32 bootaddr;
+
+ dn = of_get_cpu_node(1, NULL);
+ if (!dn) {
+ pr_err("CPU#1: missing device tree node\n");
+ return;
+ }
+ /*
+ * Determine the address from which the CPU is polling.
+ * The bootloader *does* change this property
+ */
+ ret = of_property_read_u32(dn, "cpu-release-addr", &bootaddr);
+ of_node_put(dn);
+ if (ret) {
+ pr_err("CPU#1: invalid cpu-release-addr property\n");
+ return;
+ }
+ pr_info("CPU#1: cpu-release-addr %08x\n", (u32)bootaddr);
+
+ cpu_bootaddr = ioremap(bootaddr, sizeof(bootaddr));
+ if (!cpu_bootaddr)
+ pr_err("CPU#1: cpu-release-addr map failed\n");
+}
+
+static const struct smp_operations rzn1_smp_ops __initconst = {
+ .smp_prepare_cpus = rzn1_smp_prepare_cpus,
+ .smp_boot_secondary = rzn1_smp_boot_secondary,
+};
+CPU_METHOD_OF_DECLARE(rzn1_smp, "renesas,r9a06g032-smp", &rzn1_smp_ops);
--
2.7.4
^ permalink raw reply related
* [PATCH v3 1/3] dt-bindings: cpu: Add Renesas R9A06G032 SMP enable method.
From: Michel Pollet @ 2018-05-24 10:30 UTC (permalink / raw)
To: linux-arm-kernel
In-Reply-To: <1527157834-7747-1-git-send-email-michel.pollet@bp.renesas.com>
Add a special enable method for second CA7 of the R9A06G032
Signed-off-by: Michel Pollet <michel.pollet@bp.renesas.com>
Reviewed-by: Rob Herring <robh@kernel.org>
---
Documentation/devicetree/bindings/arm/cpus.txt | 1 +
1 file changed, 1 insertion(+)
diff --git a/Documentation/devicetree/bindings/arm/cpus.txt b/Documentation/devicetree/bindings/arm/cpus.txt
index 29e1dc5..b395d107 100644
--- a/Documentation/devicetree/bindings/arm/cpus.txt
+++ b/Documentation/devicetree/bindings/arm/cpus.txt
@@ -219,6 +219,7 @@ described below.
"qcom,kpss-acc-v1"
"qcom,kpss-acc-v2"
"renesas,apmu"
+ "renesas,r9a06g032-smp"
"rockchip,rk3036-smp"
"rockchip,rk3066-smp"
"ste,dbx500-smp"
--
2.7.4
^ permalink raw reply related
* [PATCH v3 0/3] Renesas R9A06G032 SMP enabler
From: Michel Pollet @ 2018-05-24 10:30 UTC (permalink / raw)
To: linux-arm-kernel
*WARNING -- this requires the base R9A06G032 support patches
already posted
This patch series is for enabling the second CA7 of the R9A06G032.
It's based on a spin_table method, and it reuses the same binding
property as that driver.
v3:
+ Removed mentions of rz/?n1d?
+ Rebased on base patch v7
v2:
+ Added suggestions from Florian Fainelli
+ Use __pa_symbol()
+ Simplified logic in prepare_cpu()
+ Reordered the patches
+ Rebased on RZN1 Base patch v5
Michel Pollet (3):
dt-bindings: cpu: Add Renesas R9A06G032 SMP enable method.
arm: shmobile: Add the R9A06G032 SMP enabler driver
ARM: dts: Renesas R9A06G032 SMP enable method
Documentation/devicetree/bindings/arm/cpus.txt | 1 +
arch/arm/boot/dts/r9a06g032.dtsi | 2 +
arch/arm/mach-shmobile/Makefile | 1 +
arch/arm/mach-shmobile/smp-r9a06g032.c | 85 ++++++++++++++++++++++++++
4 files changed, 89 insertions(+)
create mode 100644 arch/arm/mach-shmobile/smp-r9a06g032.c
--
2.7.4
^ permalink raw reply
* [PATCHv2] arm64: Make sure permission updates happen for pmd/pud
From: Will Deacon @ 2018-05-24 10:27 UTC (permalink / raw)
To: linux-arm-kernel
In-Reply-To: <20180523184346.487-1-labbott@redhat.com>
Hi Laura,
On Wed, May 23, 2018 at 11:43:46AM -0700, Laura Abbott wrote:
> Commit 15122ee2c515 ("arm64: Enforce BBM for huge IO/VMAP mappings")
> disallowed block mappings for ioremap since that code does not honor
> break-before-make. The same APIs are also used for permission updating
> though and the extra checks prevent the permission updates from happening,
> even though this should be permitted. This results in read-only permissions
> not being fully applied. Visibly, this can occasionaly be seen as a failure
> on the built in rodata test when the test data ends up in a section or
> as an odd RW gap on the page table dump. Fix this by using
> pgattr_change_is_safe instead of p*d_present for determining if the
> change is permitted.
>
> Reported-by: Peter Robinson <pbrobinson@gmail.com>
> Fixes: 15122ee2c515 ("arm64: Enforce BBM for huge IO/VMAP mappings")
> Signed-off-by: Laura Abbott <labbott@redhat.com>
> ---
> v2: Switch to using pgattr_change_is_safe per suggestion of Will
> ---
Thanks for re-spinning so quickly. I'll queue as a fix with the relevant
tags.
Will
^ permalink raw reply
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox