TWxSoftware/twx-linux
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Merge tag 'kvm-x86-fixes-6.11-rcN' of https://github.com/kvm-x86/linux into kvm-master

Browse Source 
KVM x86 fixes for 6.11

 - Fixup missed comments from the REMOVED_SPTE=>FROZEN_SPTE rename.

 - Ensure a root is successfully loaded when pre-faulting SPTEs.

 - Grab kvm->srcu when handling KVM_SET_VCPU_EVENTS to guard against accessing
   memslots if toggling SMM happens to force a VM-Exit.

 - Emulate MSR_{FS,GS}_BASE on SVM even though interception is always disabled,
   so that KVM does the right thing if KVM's emulator encounters {RD,WR}MSR.

 - Explicitly clear BUS_LOCK_DETECT from KVM's caps on AMD, as KVM doesn't yet
   virtualize BUS_LOCK_DETECT on AMD.

 - Cleanup the help message for CONFIG_KVM_AMD_SEV, and call out that KVM now
   supports SEV-SNP too.
This commit is contained in:
Paolo Bonzini
2024-09-02 10:55:27 -04:00
parent 66155de93b 5fa9f0480c
commit 1ae99594fb
410 changed files with 4932 additions and 2795 deletions
Download Patch File Download Diff File Expand all files Collapse all files
Documentation/ABI/testing/sysfs-devices-system-cpu
+2 -1
View File
@@ -562,7 +562,8 @@ Description: Control Symmetric Multi Threading (SMT)
================ =========================================
If control status is "forceoff" or "notsupported" writes
are rejected.
are rejected. Note that enabling SMT on PowerPC skips
offline cores.
What: /sys/devices/system/cpu/cpuX/power/energy_perf_bias
Date: March 2019
Documentation/admin-guide/device-mapper/dm-crypt.rst
+8 -7
View File
@@ -162,13 +162,14 @@ iv_large_sectors
Module parameters::
max_read_size
max_write_size
Maximum size of read or write requests. When a request larger than this size
is received, dm-crypt will split the request. The splitting improves
concurrency (the split requests could be encrypted in parallel by multiple
cores), but it also causes overhead. The user should tune these parameters to
fit the actual workload.
max_read_size
max_write_size
Maximum size of read or write requests. When a request larger than this size
is received, dm-crypt will split the request. The splitting improves
concurrency (the split requests could be encrypted in parallel by multiple
cores), but it also causes overhead. The user should tune these parameters to
fit the actual workload.
Example scripts
Documentation/arch/riscv/hwprobe.rst
+22 -14
View File
@@ -239,25 +239,33 @@ The following keys are defined:
ratified in commit 98918c844281 ("Merge pull request #1217 from
riscv/zawrs") of riscv-isa-manual.
* :c:macro:`RISCV_HWPROBE_KEY_CPUPERF_0`: A bitmask that contains performance
information about the selected set of processors.
* :c:macro:`RISCV_HWPROBE_KEY_CPUPERF_0`: Deprecated. Returns similar values to
:c:macro:`RISCV_HWPROBE_KEY_MISALIGNED_SCALAR_PERF`, but the key was
mistakenly classified as a bitmask rather than a value.
* :c:macro:`RISCV_HWPROBE_MISALIGNED_UNKNOWN`: The performance of misaligned
accesses is unknown.
* :c:macro:`RISCV_HWPROBE_KEY_MISALIGNED_SCALAR_PERF`: An enum value describing
the performance of misaligned scalar native word accesses on the selected set
of processors.
* :c:macro:`RISCV_HWPROBE_MISALIGNED_EMULATED`: Misaligned accesses are
emulated via software, either in or below the kernel. These accesses are
always extremely slow.
* :c:macro:`RISCV_HWPROBE_MISALIGNED_SCALAR_UNKNOWN`: The performance of
misaligned scalar accesses is unknown.
* :c:macro:`RISCV_HWPROBE_MISALIGNED_SLOW`: Misaligned accesses are slower
than equivalent byte accesses. Misaligned accesses may be supported
directly in hardware, or trapped and emulated by software.
* :c:macro:`RISCV_HWPROBE_MISALIGNED_SCALAR_EMULATED`: Misaligned scalar
accesses are emulated via software, either in or below the kernel. These
accesses are always extremely slow.
* :c:macro:`RISCV_HWPROBE_MISALIGNED_FAST`: Misaligned accesses are faster
than equivalent byte accesses.
* :c:macro:`RISCV_HWPROBE_MISALIGNED_SCALAR_SLOW`: Misaligned scalar native
word sized accesses are slower than the equivalent quantity of byte
accesses. Misaligned accesses may be supported directly in hardware, or
trapped and emulated by software.
* :c:macro:`RISCV_HWPROBE_MISALIGNED_UNSUPPORTED`: Misaligned accesses are
not supported at all and will generate a misaligned address fault.
* :c:macro:`RISCV_HWPROBE_MISALIGNED_SCALAR_FAST`: Misaligned scalar native
word sized accesses are faster than the equivalent quantity of byte
accesses.
* :c:macro:`RISCV_HWPROBE_MISALIGNED_SCALAR_UNSUPPORTED`: Misaligned scalar
accesses are not supported at all and will generate a misaligned address
fault.
* :c:macro:`RISCV_HWPROBE_KEY_ZICBOZ_BLOCK_SIZE`: An unsigned int which
represents the size of the Zicboz block in bytes.
Documentation/devicetree/bindings/clock/qcom,dispcc-sm6350.yaml
+1 -1
View File
@@ -7,7 +7,7 @@ $schema: http://devicetree.org/meta-schemas/core.yaml#
title: Qualcomm Display Clock & Reset Controller on SM6350
maintainers:
- Konrad Dybcio <konrad.dybcio@somainline.org>
- Konrad Dybcio <konradybcio@kernel.org>
description: |
Qualcomm display clock control module provides the clocks, resets and power
Documentation/devicetree/bindings/clock/qcom,gcc-msm8994.yaml
+1 -1
View File
@@ -7,7 +7,7 @@ $schema: http://devicetree.org/meta-schemas/core.yaml#
title: Qualcomm Global Clock & Reset Controller on MSM8994
maintainers:
- Konrad Dybcio <konrad.dybcio@somainline.org>
- Konrad Dybcio <konradybcio@kernel.org>
description: |
Qualcomm global clock control module provides the clocks, resets and power
Documentation/devicetree/bindings/clock/qcom,gcc-sm6125.yaml
+1 -1
View File
@@ -7,7 +7,7 @@ $schema: http://devicetree.org/meta-schemas/core.yaml#
title: Qualcomm Global Clock & Reset Controller on SM6125
maintainers:
- Konrad Dybcio <konrad.dybcio@somainline.org>
- Konrad Dybcio <konradybcio@kernel.org>
description: |
Qualcomm global clock control module provides the clocks, resets and power
Documentation/devicetree/bindings/clock/qcom,gcc-sm6350.yaml
+1 -1
View File
@@ -7,7 +7,7 @@ $schema: http://devicetree.org/meta-schemas/core.yaml#
title: Qualcomm Global Clock & Reset Controller on SM6350
maintainers:
- Konrad Dybcio <konrad.dybcio@somainline.org>
- Konrad Dybcio <konradybcio@kernel.org>
description: |
Qualcomm global clock control module provides the clocks, resets and power
Documentation/devicetree/bindings/clock/qcom,sm6115-gpucc.yaml
+1 -1
View File
@@ -7,7 +7,7 @@ $schema: http://devicetree.org/meta-schemas/core.yaml#
title: Qualcomm Graphics Clock & Reset Controller on SM6115
maintainers:
- Konrad Dybcio <konrad.dybcio@linaro.org>
- Konrad Dybcio <konradybcio@kernel.org>
description: |
Qualcomm graphics clock control module provides clocks, resets and power
Documentation/devicetree/bindings/clock/qcom,sm6125-gpucc.yaml
+1 -1
View File
@@ -7,7 +7,7 @@ $schema: http://devicetree.org/meta-schemas/core.yaml#
title: Qualcomm Graphics Clock & Reset Controller on SM6125
maintainers:
- Konrad Dybcio <konrad.dybcio@linaro.org>
- Konrad Dybcio <konradybcio@kernel.org>
description: |
Qualcomm graphics clock control module provides clocks and power domains on
Documentation/devicetree/bindings/clock/qcom,sm6350-camcc.yaml
+1 -1
View File
@@ -7,7 +7,7 @@ $schema: http://devicetree.org/meta-schemas/core.yaml#
title: Qualcomm Camera Clock & Reset Controller on SM6350
maintainers:
- Konrad Dybcio <konrad.dybcio@linaro.org>
- Konrad Dybcio <konradybcio@kernel.org>
description: |
Qualcomm camera clock control module provides the clocks, resets and power
Documentation/devicetree/bindings/clock/qcom,sm6375-dispcc.yaml
+1 -1
View File
@@ -7,7 +7,7 @@ $schema: http://devicetree.org/meta-schemas/core.yaml#
title: Qualcomm Display Clock & Reset Controller on SM6375
maintainers:
- Konrad Dybcio <konrad.dybcio@linaro.org>
- Konrad Dybcio <konradybcio@kernel.org>
description: |
Qualcomm display clock control module provides the clocks, resets and power
Documentation/devicetree/bindings/clock/qcom,sm6375-gcc.yaml
+1 -1
View File
@@ -7,7 +7,7 @@ $schema: http://devicetree.org/meta-schemas/core.yaml#
title: Qualcomm Global Clock & Reset Controller on SM6375
maintainers:
- Konrad Dybcio <konrad.dybcio@somainline.org>
- Konrad Dybcio <konradybcio@kernel.org>
description: |
Qualcomm global clock control module provides the clocks, resets and power
Documentation/devicetree/bindings/clock/qcom,sm6375-gpucc.yaml
+1 -1
View File
@@ -7,7 +7,7 @@ $schema: http://devicetree.org/meta-schemas/core.yaml#
title: Qualcomm Graphics Clock & Reset Controller on SM6375
maintainers:
- Konrad Dybcio <konrad.dybcio@linaro.org>
- Konrad Dybcio <konradybcio@kernel.org>
description: |
Qualcomm graphics clock control module provides clocks, resets and power
Documentation/devicetree/bindings/clock/qcom,sm8350-videocc.yaml
+1 -1
View File
@@ -7,7 +7,7 @@ $schema: http://devicetree.org/meta-schemas/core.yaml#
title: Qualcomm SM8350 Video Clock & Reset Controller
maintainers:
- Konrad Dybcio <konrad.dybcio@linaro.org>
- Konrad Dybcio <konradybcio@kernel.org>
description: |
Qualcomm video clock control module provides the clocks, resets and power
Documentation/devicetree/bindings/clock/qcom,sm8450-gpucc.yaml
+1 -1
View File
@@ -7,7 +7,7 @@ $schema: http://devicetree.org/meta-schemas/core.yaml#
title: Qualcomm Graphics Clock & Reset Controller on SM8450
maintainers:
- Konrad Dybcio <konrad.dybcio@linaro.org>
- Konrad Dybcio <konradybcio@kernel.org>
description: |
Qualcomm graphics clock control module provides the clocks, resets and power
Documentation/devicetree/bindings/display/msm/qcom,sm6375-mdss.yaml
+1 -1
View File
@@ -7,7 +7,7 @@ $schema: http://devicetree.org/meta-schemas/core.yaml#
title: Qualcomm SM6375 Display MDSS
maintainers:
- Konrad Dybcio <konrad.dybcio@linaro.org>
- Konrad Dybcio <konradybcio@kernel.org>
description:
SM6375 MSM Mobile Display Subsystem (MDSS), which encapsulates sub-blocks
Documentation/devicetree/bindings/display/panel/asus,z00t-tm5p5-nt35596.yaml
+1 -1
View File
@@ -7,7 +7,7 @@ $schema: http://devicetree.org/meta-schemas/core.yaml#
title: ASUS Z00T TM5P5 NT35596 5.5" 1080×1920 LCD Panel
maintainers:
- Konrad Dybcio <konradybcio@gmail.com>
- Konrad Dybcio <konradybcio@kernel.org>
description: |+
This panel seems to only be found in the Asus Z00T
Documentation/devicetree/bindings/display/panel/samsung,atna33xc20.yaml
+6 -6
View File
@@ -18,12 +18,12 @@ properties:
# Samsung 13.3" FHD (1920x1080 pixels) eDP AMOLED panel
- const: samsung,atna33xc20
- items:
- enum:
# Samsung 14.5" WQXGA+ (2880x1800 pixels) eDP AMOLED panel
- samsung,atna45af01
# Samsung 14.5" 3K (2944x1840 pixels) eDP AMOLED panel
- samsung,atna45dc02
- const: samsung,atna33xc20
- enum:
# Samsung 14.5" WQXGA+ (2880x1800 pixels) eDP AMOLED panel
- samsung,atna45af01
# Samsung 14.5" 3K (2944x1840 pixels) eDP AMOLED panel
- samsung,atna45dc02
- const: samsung,atna33xc20
enable-gpios: true
port: true
Documentation/devicetree/bindings/display/panel/sony,td4353-jdi.yaml
+1 -1
View File
@@ -7,7 +7,7 @@ $schema: http://devicetree.org/meta-schemas/core.yaml#
title: Sony TD4353 JDI 5 / 5.7" 2160x1080 MIPI-DSI Panel
maintainers:
- Konrad Dybcio <konrad.dybcio@somainline.org>
- Konrad Dybcio <konradybcio@kernel.org>
description: |
The Sony TD4353 JDI is a 5 (XZ2c) / 5.7 (XZ2) inch 2160x1080
Documentation/devicetree/bindings/eeprom/at25.yaml
+1
View File
@@ -28,6 +28,7 @@ properties:
- anvo,anv32e61w
- atmel,at25256B
- fujitsu,mb85rs1mt
- fujitsu,mb85rs256
- fujitsu,mb85rs64
- microchip,at25160bn
- microchip,25lc040
Documentation/devicetree/bindings/interconnect/qcom,sc7280-rpmh.yaml
+1 -1
View File
@@ -8,7 +8,7 @@ title: Qualcomm RPMh Network-On-Chip Interconnect on SC7280
maintainers:
- Bjorn Andersson <andersson@kernel.org>
- Konrad Dybcio <konrad.dybcio@linaro.org>
- Konrad Dybcio <konradybcio@kernel.org>
description: |
RPMh interconnect providers support system bandwidth requirements through
Documentation/devicetree/bindings/interconnect/qcom,sc8280xp-rpmh.yaml
+1 -1
View File
@@ -8,7 +8,7 @@ title: Qualcomm RPMh Network-On-Chip Interconnect on SC8280XP
maintainers:
- Bjorn Andersson <andersson@kernel.org>
- Konrad Dybcio <konrad.dybcio@linaro.org>
- Konrad Dybcio <konradybcio@kernel.org>
description: |
RPMh interconnect providers support system bandwidth requirements through
Documentation/devicetree/bindings/interconnect/qcom,sm8450-rpmh.yaml
+1 -1
View File
@@ -8,7 +8,7 @@ title: Qualcomm RPMh Network-On-Chip Interconnect on SM8450
maintainers:
- Bjorn Andersson <andersson@kernel.org>
- Konrad Dybcio <konrad.dybcio@linaro.org>
- Konrad Dybcio <konradybcio@kernel.org>
description: |
RPMh interconnect providers support system bandwidth requirements through
Documentation/devicetree/bindings/iommu/qcom,iommu.yaml
+1 -1
View File
@@ -7,7 +7,7 @@ $schema: http://devicetree.org/meta-schemas/core.yaml#
title: Qualcomm Technologies legacy IOMMU implementations
maintainers:
- Konrad Dybcio <konrad.dybcio@linaro.org>
- Konrad Dybcio <konradybcio@kernel.org>
description: |
Qualcomm "B" family devices which are not compatible with arm-smmu have
Documentation/devicetree/bindings/net/fsl,qoriq-mc-dpmac.yaml
+4
View File
@@ -38,6 +38,10 @@ properties:
managed: true
phys:
description: A reference to the SerDes lane(s)
maxItems: 1
required:
- reg
Documentation/devicetree/bindings/pinctrl/qcom,mdm9607-tlmm.yaml
+1 -1
View File
@@ -7,7 +7,7 @@ $schema: http://devicetree.org/meta-schemas/core.yaml#
title: Qualcomm Technologies, Inc. MDM9607 TLMM block
maintainers:
- Konrad Dybcio <konrad.dybcio@somainline.org>
- Konrad Dybcio <konradybcio@kernel.org>
description:
Top Level Mode Multiplexer pin controller in Qualcomm MDM9607 SoC.
Documentation/devicetree/bindings/pinctrl/qcom,sm6350-tlmm.yaml
+1 -1
View File
@@ -7,7 +7,7 @@ $schema: http://devicetree.org/meta-schemas/core.yaml#
title: Qualcomm Technologies, Inc. SM6350 TLMM block
maintainers:
- Konrad Dybcio <konrad.dybcio@somainline.org>
- Konrad Dybcio <konradybcio@kernel.org>
description:
Top Level Mode Multiplexer pin controller in Qualcomm SM6350 SoC.
Documentation/devicetree/bindings/pinctrl/qcom,sm6375-tlmm.yaml
+1 -1
View File
@@ -7,7 +7,7 @@ $schema: http://devicetree.org/meta-schemas/core.yaml#
title: Qualcomm Technologies, Inc. SM6375 TLMM block
maintainers:
- Konrad Dybcio <konrad.dybcio@somainline.org>
- Konrad Dybcio <konradybcio@kernel.org>
description:
Top Level Mode Multiplexer pin controller in Qualcomm SM6375 SoC.
Documentation/devicetree/bindings/remoteproc/qcom,rpm-proc.yaml
+1 -1
View File
@@ -8,7 +8,7 @@ title: Qualcomm Resource Power Manager (RPM) Processor/Subsystem
maintainers:
- Bjorn Andersson <andersson@kernel.org>
- Konrad Dybcio <konrad.dybcio@linaro.org>
- Konrad Dybcio <konradybcio@kernel.org>
- Stephan Gerhold <stephan@gerhold.net>
description: |
Documentation/devicetree/bindings/soc/qcom/qcom,rpm-master-stats.yaml
+1 -1
View File
@@ -7,7 +7,7 @@ $schema: http://devicetree.org/meta-schemas/core.yaml#
title: Qualcomm Technologies, Inc. (QTI) RPM Master Stats
maintainers:
- Konrad Dybcio <konrad.dybcio@linaro.org>
- Konrad Dybcio <konradybcio@kernel.org>
description: |
The Qualcomm RPM (Resource Power Manager) architecture includes a concept
Documentation/filesystems/caching/fscache.rst
+4 -4
View File
@@ -318,10 +318,10 @@ where the columns are:
Debugging
=========
If CONFIG_FSCACHE_DEBUG is enabled, the FS-Cache facility can have runtime
debugging enabled by adjusting the value in::
If CONFIG_NETFS_DEBUG is enabled, the FS-Cache facility and NETFS support can
have runtime debugging enabled by adjusting the value in::
/sys/module/fscache/parameters/debug
/sys/module/netfs/parameters/debug
This is a bitmask of debugging streams to enable:
@@ -343,6 +343,6 @@ This is a bitmask of debugging streams to enable:
The appropriate set of values should be OR'd together and the result written to
the control file. For example::
echo $((1|8|512)) >/sys/module/fscache/parameters/debug
echo $((1|8|512)) >/sys/module/netfs/parameters/debug
will turn on all function entry debugging.
Makefile
+2 -2
View File
@@ -2,7 +2,7 @@
VERSION = 6
PATCHLEVEL = 11
SUBLEVEL = 0
EXTRAVERSION = -rc3
EXTRAVERSION = -rc4
NAME = Baby Opossum Posse
# *DOCUMENTATION*
@@ -1963,7 +1963,7 @@ tags TAGS cscope gtags: FORCE
# Protocol).
PHONY += rust-analyzer
rust-analyzer:
$(Q)$(CONFIG_SHELL) $(srctree)/scripts/rust_is_available.sh
+$(Q)$(CONFIG_SHELL) $(srctree)/scripts/rust_is_available.sh
$(Q)$(MAKE) $(build)=rust $@
# Script to generate missing namespace dependencies
arch/arm/mach-rpc/ecard.c
+1 -1
View File
@@ -1109,7 +1109,7 @@ void ecard_remove_driver(struct ecard_driver *drv)
driver_unregister(&drv->drv);
}
static int ecard_match(struct device *_dev, struct device_driver *_drv)
static int ecard_match(struct device *_dev, const struct device_driver *_drv)
{
struct expansion_card *ec = ECARD_DEV(_dev);
struct ecard_driver *drv = ECARD_DRV(_drv);
arch/arm64/include/asm/uaccess.h
+1 -1
View File
@@ -188,7 +188,7 @@ static inline void __user *__uaccess_mask_ptr(const void __user *ptr)
#define __get_mem_asm(load, reg, x, addr, label, type) \
asm_goto_output( \
"1: " load " " reg "0, [%1]\n" \
_ASM_EXTABLE_##type##ACCESS_ERR(1b, %l2, %w0) \
_ASM_EXTABLE_##type##ACCESS(1b, %l2) \
: "=r" (x) \
: "r" (addr) : : label)
#else
arch/arm64/kernel/acpi_numa.c
+1 -1
View File
@@ -27,7 +27,7 @@
#include <asm/numa.h>
static int acpi_early_node_map[NR_CPUS] __initdata = { NUMA_NO_NODE };
static int acpi_early_node_map[NR_CPUS] __initdata = { [0 ... NR_CPUS - 1] = NUMA_NO_NODE };
int __init acpi_numa_get_nid(unsigned int cpu)
{
arch/arm64/kernel/setup.c
-3
View File
@@ -355,9 +355,6 @@ void __init __no_sanitize_address setup_arch(char **cmdline_p)
smp_init_cpus();
smp_build_mpidr_hash();
/* Init percpu seeds for random tags after cpus are set up. */
kasan_init_sw_tags();
#ifdef CONFIG_ARM64_SW_TTBR0_PAN
/*
* Make sure init_thread_info.ttbr0 always generates translation
arch/arm64/kernel/smp.c
+2
View File
@@ -467,6 +467,8 @@ void __init smp_prepare_boot_cpu(void)
init_gic_priority_masking();
kasan_init_hw_tags();
/* Init percpu seeds for random tags after cpus are set up. */
kasan_init_sw_tags();
}
/*
arch/mips/sgi-ip22/ip22-gio.c
+1 -1
View File
@@ -111,7 +111,7 @@ void gio_device_unregister(struct gio_device *giodev)
}
EXPORT_SYMBOL_GPL(gio_device_unregister);
static int gio_bus_match(struct device *dev, struct device_driver *drv)
static int gio_bus_match(struct device *dev, const struct device_driver *drv)
{
struct gio_device *gio_dev = to_gio_device(dev);
struct gio_driver *gio_drv = to_gio_driver(drv);
arch/powerpc/include/asm/topology.h
+13
View File
@@ -145,6 +145,7 @@ static inline int cpu_to_coregroup_id(int cpu)
#ifdef CONFIG_HOTPLUG_SMT
#include <linux/cpu_smt.h>
#include <linux/cpumask.h>
#include <asm/cputhreads.h>
static inline bool topology_is_primary_thread(unsigned int cpu)
@@ -156,6 +157,18 @@ static inline bool topology_smt_thread_allowed(unsigned int cpu)
{
return cpu_thread_in_core(cpu) < cpu_smt_num_threads;
}
#define topology_is_core_online topology_is_core_online
static inline bool topology_is_core_online(unsigned int cpu)
{
int i, first_cpu = cpu_first_thread_sibling(cpu);
for (i = first_cpu; i < first_cpu + threads_per_core; ++i) {
if (cpu_online(i))
return true;
}
return false;
}
#endif
#endif /* __KERNEL__ */
arch/powerpc/kernel/setup-common.c
+1
View File
@@ -959,6 +959,7 @@ void __init setup_arch(char **cmdline_p)
mem_topology_setup();
/* Set max_mapnr before paging_init() */
set_max_mapnr(max_pfn);
high_memory = (void *)__va(max_low_pfn * PAGE_SIZE);
/*
* Release secondary cpus out of their spinloops at 0x60 now that
arch/powerpc/mm/init-common.c
+2 -2
View File
@@ -73,7 +73,7 @@ void setup_kup(void)
#define CTOR(shift) static void ctor_##shift(void *addr) \
{ \
memset(addr, 0, sizeof(void *) << (shift)); \
memset(addr, 0, sizeof(pgd_t) << (shift)); \
}
CTOR(0); CTOR(1); CTOR(2); CTOR(3); CTOR(4); CTOR(5); CTOR(6); CTOR(7);
@@ -117,7 +117,7 @@ EXPORT_SYMBOL_GPL(pgtable_cache); /* used by kvm_hv module */
void pgtable_cache_add(unsigned int shift)
{
char *name;
unsigned long table_size = sizeof(void *) << shift;
unsigned long table_size = sizeof(pgd_t) << shift;
unsigned long align = table_size;
/* When batching pgtable pointers for RCU freeing, we store
arch/powerpc/mm/mem.c
-2
View File
@@ -290,8 +290,6 @@ void __init mem_init(void)
swiotlb_init(ppc_swiotlb_enable, ppc_swiotlb_flags);
#endif
high_memory = (void *) __va(max_low_pfn * PAGE_SIZE);
kasan_late_init();
memblock_free_all();
arch/riscv/include/asm/hwprobe.h
+1 -1
View File
@@ -8,7 +8,7 @@
#include <uapi/asm/hwprobe.h>
#define RISCV_HWPROBE_MAX_KEY 8
#define RISCV_HWPROBE_MAX_KEY 9
static inline bool riscv_hwprobe_key_is_valid(__s64 key)
{
arch/riscv/include/uapi/asm/hwprobe.h
+6
View File
@@ -82,6 +82,12 @@ struct riscv_hwprobe {
#define RISCV_HWPROBE_KEY_ZICBOZ_BLOCK_SIZE 6
#define RISCV_HWPROBE_KEY_HIGHEST_VIRT_ADDRESS 7
#define RISCV_HWPROBE_KEY_TIME_CSR_FREQ 8
#define RISCV_HWPROBE_KEY_MISALIGNED_SCALAR_PERF 9
#define RISCV_HWPROBE_MISALIGNED_SCALAR_UNKNOWN 0
#define RISCV_HWPROBE_MISALIGNED_SCALAR_EMULATED 1
#define RISCV_HWPROBE_MISALIGNED_SCALAR_SLOW 2
#define RISCV_HWPROBE_MISALIGNED_SCALAR_FAST 3
#define RISCV_HWPROBE_MISALIGNED_SCALAR_UNSUPPORTED 4
/* Increase RISCV_HWPROBE_MAX_KEY when adding items. */
/* Flags */
arch/riscv/kernel/acpi_numa.c
+1 -1
View File
@@ -28,7 +28,7 @@
#include <asm/numa.h>
static int acpi_early_node_map[NR_CPUS] __initdata = { NUMA_NO_NODE };
static int acpi_early_node_map[NR_CPUS] __initdata = { [0 ... NR_CPUS - 1] = NUMA_NO_NODE };
int __init acpi_numa_get_nid(unsigned int cpu)
{
arch/riscv/kernel/patch.c
+4
View File
@@ -205,6 +205,8 @@ int patch_text_set_nosync(void *addr, u8 c, size_t len)
int ret;
ret = patch_insn_set(addr, c, len);
if (!ret)
flush_icache_range((uintptr_t)addr, (uintptr_t)addr + len);
return ret;
}
@@ -239,6 +241,8 @@ int patch_text_nosync(void *addr, const void *insns, size_t len)
int ret;
ret = patch_insn_write(addr, insns, len);
if (!ret)
flush_icache_range((uintptr_t)addr, (uintptr_t)addr + len);
return ret;
}
arch/riscv/kernel/sys_hwprobe.c
+6 -5
View File
@@ -178,13 +178,13 @@ static u64 hwprobe_misaligned(const struct cpumask *cpus)
perf = this_perf;
if (perf != this_perf) {
perf = RISCV_HWPROBE_MISALIGNED_UNKNOWN;
perf = RISCV_HWPROBE_MISALIGNED_SCALAR_UNKNOWN;
break;
}
}
if (perf == -1ULL)
return RISCV_HWPROBE_MISALIGNED_UNKNOWN;
return RISCV_HWPROBE_MISALIGNED_SCALAR_UNKNOWN;
return perf;
}
@@ -192,12 +192,12 @@ static u64 hwprobe_misaligned(const struct cpumask *cpus)
static u64 hwprobe_misaligned(const struct cpumask *cpus)
{
if (IS_ENABLED(CONFIG_RISCV_EFFICIENT_UNALIGNED_ACCESS))
return RISCV_HWPROBE_MISALIGNED_FAST;
return RISCV_HWPROBE_MISALIGNED_SCALAR_FAST;
if (IS_ENABLED(CONFIG_RISCV_EMULATED_UNALIGNED_ACCESS) && unaligned_ctl_available())
return RISCV_HWPROBE_MISALIGNED_EMULATED;
return RISCV_HWPROBE_MISALIGNED_SCALAR_EMULATED;
return RISCV_HWPROBE_MISALIGNED_SLOW;
return RISCV_HWPROBE_MISALIGNED_SCALAR_SLOW;
}
#endif
@@ -225,6 +225,7 @@ static void hwprobe_one_pair(struct riscv_hwprobe *pair,
break;
case RISCV_HWPROBE_KEY_CPUPERF_0:
case RISCV_HWPROBE_KEY_MISALIGNED_SCALAR_PERF:
pair->value = hwprobe_misaligned(cpus);
break;
arch/riscv/kernel/traps.c
+2 -2
View File
@@ -319,6 +319,7 @@ void do_trap_ecall_u(struct pt_regs *regs)
regs->epc += 4;
regs->orig_a0 = regs->a0;
regs->a0 = -ENOSYS;
riscv_v_vstate_discard(regs);
@@ -328,8 +329,7 @@ void do_trap_ecall_u(struct pt_regs *regs)
if (syscall >= 0 && syscall < NR_syscalls)
syscall_handler(regs, syscall);
else if (syscall != -1)
regs->a0 = -ENOSYS;
/*
* Ultimately, this value will get limited by KSTACK_OFFSET_MAX(),
* so the maximum stack offset is 1k bytes (10 bits).
arch/riscv/kernel/traps_misaligned.c
+3 -3
View File
@@ -338,7 +338,7 @@ int handle_misaligned_load(struct pt_regs *regs)
perf_sw_event(PERF_COUNT_SW_ALIGNMENT_FAULTS, 1, regs, addr);
#ifdef CONFIG_RISCV_PROBE_UNALIGNED_ACCESS
*this_cpu_ptr(&misaligned_access_speed) = RISCV_HWPROBE_MISALIGNED_EMULATED;
*this_cpu_ptr(&misaligned_access_speed) = RISCV_HWPROBE_MISALIGNED_SCALAR_EMULATED;
#endif
if (!unaligned_enabled)
@@ -532,13 +532,13 @@ static bool check_unaligned_access_emulated(int cpu)
unsigned long tmp_var, tmp_val;
bool misaligned_emu_detected;
*mas_ptr = RISCV_HWPROBE_MISALIGNED_UNKNOWN;
*mas_ptr = RISCV_HWPROBE_MISALIGNED_SCALAR_UNKNOWN;
__asm__ __volatile__ (
" "REG_L" %[tmp], 1(%[ptr])\n"
: [tmp] "=r" (tmp_val) : [ptr] "r" (&tmp_var) : "memory");
misaligned_emu_detected = (*mas_ptr == RISCV_HWPROBE_MISALIGNED_EMULATED);
misaligned_emu_detected = (*mas_ptr == RISCV_HWPROBE_MISALIGNED_SCALAR_EMULATED);
/*
* If unaligned_ctl is already set, this means that we detected that all
* CPUS uses emulated misaligned access at boot time. If that changed
arch/riscv/kernel/unaligned_access_speed.c
+6 -6
View File
@@ -34,9 +34,9 @@ static int check_unaligned_access(void *param)
struct page *page = param;
void *dst;
void *src;
long speed = RISCV_HWPROBE_MISALIGNED_SLOW;
long speed = RISCV_HWPROBE_MISALIGNED_SCALAR_SLOW;
if (per_cpu(misaligned_access_speed, cpu) != RISCV_HWPROBE_MISALIGNED_UNKNOWN)
if (per_cpu(misaligned_access_speed, cpu) != RISCV_HWPROBE_MISALIGNED_SCALAR_UNKNOWN)
return 0;
/* Make an unaligned destination buffer. */
@@ -95,14 +95,14 @@ static int check_unaligned_access(void *param)
}
if (word_cycles < byte_cycles)
speed = RISCV_HWPROBE_MISALIGNED_FAST;
speed = RISCV_HWPROBE_MISALIGNED_SCALAR_FAST;
ratio = div_u64((byte_cycles * 100), word_cycles);
pr_info("cpu%d: Ratio of byte access time to unaligned word access is %d.%02d, unaligned accesses are %s\n",
cpu,
ratio / 100,
ratio % 100,
(speed == RISCV_HWPROBE_MISALIGNED_FAST) ? "fast" : "slow");
(speed == RISCV_HWPROBE_MISALIGNED_SCALAR_FAST) ? "fast" : "slow");
per_cpu(misaligned_access_speed, cpu) = speed;
@@ -110,7 +110,7 @@ static int check_unaligned_access(void *param)
* Set the value of fast_misaligned_access of a CPU. These operations
* are atomic to avoid race conditions.
*/
if (speed == RISCV_HWPROBE_MISALIGNED_FAST)
if (speed == RISCV_HWPROBE_MISALIGNED_SCALAR_FAST)
cpumask_set_cpu(cpu, &fast_misaligned_access);
else
cpumask_clear_cpu(cpu, &fast_misaligned_access);
@@ -188,7 +188,7 @@ static int riscv_online_cpu(unsigned int cpu)
static struct page *buf;
/* We are already set since the last check */
if (per_cpu(misaligned_access_speed, cpu) != RISCV_HWPROBE_MISALIGNED_UNKNOWN)
if (per_cpu(misaligned_access_speed, cpu) != RISCV_HWPROBE_MISALIGNED_SCALAR_UNKNOWN)
goto exit;
buf = alloc_pages(GFP_KERNEL, MISALIGNED_BUFFER_ORDER);
arch/riscv/kernel/vendor_extensions.c
+1 -1
View File
@@ -38,7 +38,7 @@ bool __riscv_isa_vendor_extension_available(int cpu, unsigned long vendor, unsig
#ifdef CONFIG_RISCV_ISA_VENDOR_EXT_ANDES
case ANDES_VENDOR_ID:
bmap = &riscv_isa_vendor_ext_list_andes.all_harts_isa_bitmap;
cpu_bmap = &riscv_isa_vendor_ext_list_andes.per_hart_isa_bitmap[cpu];
cpu_bmap = riscv_isa_vendor_ext_list_andes.per_hart_isa_bitmap;
break;
#endif
default:
arch/riscv/mm/init.c
+2 -2
View File
@@ -927,7 +927,7 @@ static void __init create_kernel_page_table(pgd_t *pgdir,
PMD_SIZE, PAGE_KERNEL_EXEC);
/* Map the data in RAM */
end_va = kernel_map.virt_addr + XIP_OFFSET + kernel_map.size;
end_va = kernel_map.virt_addr + kernel_map.size;
for (va = kernel_map.virt_addr + XIP_OFFSET; va < end_va; va += PMD_SIZE)
create_pgd_mapping(pgdir, va,
kernel_map.phys_addr + (va - (kernel_map.virt_addr + XIP_OFFSET)),
@@ -1096,7 +1096,7 @@ asmlinkage void __init setup_vm(uintptr_t dtb_pa)
phys_ram_base = CONFIG_PHYS_RAM_BASE;
kernel_map.phys_addr = (uintptr_t)CONFIG_PHYS_RAM_BASE;
kernel_map.size = (uintptr_t)(&_end) - (uintptr_t)(&_sdata);
kernel_map.size = (uintptr_t)(&_end) - (uintptr_t)(&_start);
kernel_map.va_kernel_xip_pa_offset = kernel_map.virt_addr - kernel_map.xiprom;
#else
arch/x86/kvm/Kconfig
+4 -2
View File
@@ -144,8 +144,10 @@ config KVM_AMD_SEV
select HAVE_KVM_ARCH_GMEM_PREPARE
select HAVE_KVM_ARCH_GMEM_INVALIDATE
help
Provides support for launching Encrypted VMs (SEV) and Encrypted VMs
with Encrypted State (SEV-ES) on AMD processors.
Provides support for launching encrypted VMs which use Secure
Encrypted Virtualization (SEV), Secure Encrypted Virtualization with
Encrypted State (SEV-ES), and Secure Encrypted Virtualization with
Secure Nested Paging (SEV-SNP) technologies on AMD processors.
config KVM_SMM
bool "System Management Mode emulation"
arch/x86/kvm/mmu/mmu.c
+3 -1
View File
@@ -4750,7 +4750,9 @@ long kvm_arch_vcpu_pre_fault_memory(struct kvm_vcpu *vcpu,
* reload is efficient when called repeatedly, so we can do it on
* every iteration.
*/
kvm_mmu_reload(vcpu);
r = kvm_mmu_reload(vcpu);
if (r)
return r;
if (kvm_arch_has_private_mem(vcpu->kvm) &&
kvm_mem_is_private(vcpu->kvm, gpa_to_gfn(range->gpa)))
arch/x86/kvm/mmu/spte.c
+3 -3
View File
@@ -391,9 +391,9 @@ void kvm_mmu_set_mmio_spte_mask(u64 mmio_value, u64 mmio_mask, u64 access_mask)
mmio_value = 0;
/*
* The masked MMIO value must obviously match itself and a removed SPTE
* must not get a false positive. Removed SPTEs and MMIO SPTEs should
* never collide as MMIO must set some RWX bits, and removed SPTEs must
* The masked MMIO value must obviously match itself and a frozen SPTE
* must not get a false positive. Frozen SPTEs and MMIO SPTEs should
* never collide as MMIO must set some RWX bits, and frozen SPTEs must
* not set any RWX bits.
*/
if (WARN_ON((mmio_value & mmio_mask) != mmio_value) ||
arch/x86/kvm/mmu/spte.h
+1 -1
View File
@@ -214,7 +214,7 @@ extern u64 __read_mostly shadow_nonpresent_or_rsvd_mask;
*/
#define FROZEN_SPTE (SHADOW_NONPRESENT_VALUE | 0x5a0ULL)
/* Removed SPTEs must not be misconstrued as shadow present PTEs. */
/* Frozen SPTEs must not be misconstrued as shadow present PTEs. */
static_assert(!(FROZEN_SPTE & SPTE_MMU_PRESENT_MASK));
static inline bool is_frozen_spte(u64 spte)
arch/x86/kvm/mmu/tdp_mmu.c
+4 -4
View File
@@ -359,10 +359,10 @@ static void handle_removed_pt(struct kvm *kvm, tdp_ptep_t pt, bool shared)
/*
* Set the SPTE to a nonpresent value that other
* threads will not overwrite. If the SPTE was
* already marked as removed then another thread
* already marked as frozen then another thread
* handling a page fault could overwrite it, so
* set the SPTE until it is set from some other
* value to the removed SPTE value.
* value to the frozen SPTE value.
*/
for (;;) {
old_spte = kvm_tdp_mmu_write_spte_atomic(sptep, FROZEN_SPTE);
@@ -536,8 +536,8 @@ static inline int __must_check __tdp_mmu_set_spte_atomic(struct tdp_iter *iter,
u64 *sptep = rcu_dereference(iter->sptep);
/*
* The caller is responsible for ensuring the old SPTE is not a REMOVED
* SPTE. KVM should never attempt to zap or manipulate a REMOVED SPTE,
* The caller is responsible for ensuring the old SPTE is not a FROZEN
* SPTE. KVM should never attempt to zap or manipulate a FROZEN SPTE,
* and pre-checking before inserting a new SPTE is advantageous as it
* avoids unnecessary work.
*/
arch/x86/kvm/svm/svm.c
+15
View File
@@ -2876,6 +2876,12 @@ static int svm_get_msr(struct kvm_vcpu *vcpu, struct msr_data *msr_info)
case MSR_CSTAR:
msr_info->data = svm->vmcb01.ptr->save.cstar;
break;
case MSR_GS_BASE:
msr_info->data = svm->vmcb01.ptr->save.gs.base;
break;
case MSR_FS_BASE:
msr_info->data = svm->vmcb01.ptr->save.fs.base;
break;
case MSR_KERNEL_GS_BASE:
msr_info->data = svm->vmcb01.ptr->save.kernel_gs_base;
break;
@@ -3101,6 +3107,12 @@ static int svm_set_msr(struct kvm_vcpu *vcpu, struct msr_data *msr)
case MSR_CSTAR:
svm->vmcb01.ptr->save.cstar = data;
break;
case MSR_GS_BASE:
svm->vmcb01.ptr->save.gs.base = data;
break;
case MSR_FS_BASE:
svm->vmcb01.ptr->save.fs.base = data;
break;
case MSR_KERNEL_GS_BASE:
svm->vmcb01.ptr->save.kernel_gs_base = data;
break;
@@ -5224,6 +5236,9 @@ static __init void svm_set_cpu_caps(void)
/* CPUID 0x8000001F (SME/SEV features) */
sev_set_cpu_caps();
/* Don't advertise Bus Lock Detect to guest if SVM support is absent */
kvm_cpu_cap_clear(X86_FEATURE_BUS_LOCK_DETECT);
}
static __init int svm_hardware_setup(void)
arch/x86/kvm/x86.c
+2
View File
@@ -6040,7 +6040,9 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
if (copy_from_user(&events, argp, sizeof(struct kvm_vcpu_events)))
break;
kvm_vcpu_srcu_read_lock(vcpu);
r = kvm_vcpu_ioctl_x86_set_vcpu_events(vcpu, &events);
kvm_vcpu_srcu_read_unlock(vcpu);
break;
}
case KVM_GET_DEBUGREGS: {
block/blk-mq-tag.c
+3 -2
View File
@@ -38,6 +38,7 @@ static void blk_mq_update_wake_batch(struct blk_mq_tags *tags,
void __blk_mq_tag_busy(struct blk_mq_hw_ctx *hctx)
{
unsigned int users;
unsigned long flags;
struct blk_mq_tags *tags = hctx->tags;
/*
@@ -56,11 +57,11 @@ void __blk_mq_tag_busy(struct blk_mq_hw_ctx *hctx)
return;
}
spin_lock_irq(&tags->lock);
spin_lock_irqsave(&tags->lock, flags);
users = tags->active_queues + 1;
WRITE_ONCE(tags->active_queues, users);
blk_mq_update_wake_batch(tags, users);
spin_unlock_irq(&tags->lock);
spin_unlock_irqrestore(&tags->lock, flags);
}
/*
drivers/acpi/acpica/acevents.h
+1 -5
View File
@@ -188,13 +188,9 @@ acpi_ev_detach_region(union acpi_operand_object *region_obj,
u8 acpi_ns_is_locked);
void
acpi_ev_execute_reg_methods(struct acpi_namespace_node *node,
acpi_ev_execute_reg_methods(struct acpi_namespace_node *node, u32 max_depth,
acpi_adr_space_type space_id, u32 function);
void
acpi_ev_execute_orphan_reg_method(struct acpi_namespace_node *node,
acpi_adr_space_type space_id);
acpi_status
acpi_ev_execute_reg_method(union acpi_operand_object *region_obj, u32 function);
drivers/acpi/acpica/evregion.c
+9 -3
View File
@@ -20,6 +20,10 @@ extern u8 acpi_gbl_default_address_spaces[];
/* Local prototypes */
static void
acpi_ev_execute_orphan_reg_method(struct acpi_namespace_node *device_node,
acpi_adr_space_type space_id);
static acpi_status
acpi_ev_reg_run(acpi_handle obj_handle,
u32 level, void *context, void **return_value);
@@ -61,6 +65,7 @@ acpi_status acpi_ev_initialize_op_regions(void)
acpi_gbl_default_address_spaces
[i])) {
acpi_ev_execute_reg_methods(acpi_gbl_root_node,
ACPI_UINT32_MAX,
acpi_gbl_default_address_spaces
[i], ACPI_REG_CONNECT);
}
@@ -668,6 +673,7 @@ cleanup1:
* FUNCTION: acpi_ev_execute_reg_methods
*
* PARAMETERS: node - Namespace node for the device
* max_depth - Depth to which search for _REG
* space_id - The address space ID
* function - Passed to _REG: On (1) or Off (0)
*
@@ -679,7 +685,7 @@ cleanup1:
******************************************************************************/
void
acpi_ev_execute_reg_methods(struct acpi_namespace_node *node,
acpi_ev_execute_reg_methods(struct acpi_namespace_node *node, u32 max_depth,
acpi_adr_space_type space_id, u32 function)
{
struct acpi_reg_walk_info info;
@@ -713,7 +719,7 @@ acpi_ev_execute_reg_methods(struct acpi_namespace_node *node,
* regions and _REG methods. (i.e. handlers must be installed for all
* regions of this Space ID before we can run any _REG methods)
*/
(void)acpi_ns_walk_namespace(ACPI_TYPE_ANY, node, ACPI_UINT32_MAX,
(void)acpi_ns_walk_namespace(ACPI_TYPE_ANY, node, max_depth,
ACPI_NS_WALK_UNLOCK, acpi_ev_reg_run, NULL,
&info, NULL);
@@ -814,7 +820,7 @@ acpi_ev_reg_run(acpi_handle obj_handle,
*
******************************************************************************/
void
static void
acpi_ev_execute_orphan_reg_method(struct acpi_namespace_node *device_node,
acpi_adr_space_type space_id)
{
drivers/acpi/acpica/evxfregn.c
+7 -57
View File
@@ -85,7 +85,8 @@ acpi_install_address_space_handler_internal(acpi_handle device,
/* Run all _REG methods for this address space */
if (run_reg) {
acpi_ev_execute_reg_methods(node, space_id, ACPI_REG_CONNECT);
acpi_ev_execute_reg_methods(node, ACPI_UINT32_MAX, space_id,
ACPI_REG_CONNECT);
}
unlock_and_exit:
@@ -263,6 +264,7 @@ ACPI_EXPORT_SYMBOL(acpi_remove_address_space_handler)
* FUNCTION: acpi_execute_reg_methods
*
* PARAMETERS: device - Handle for the device
* max_depth - Depth to which search for _REG
* space_id - The address space ID
*
* RETURN: Status
@@ -271,7 +273,8 @@ ACPI_EXPORT_SYMBOL(acpi_remove_address_space_handler)
*
******************************************************************************/
acpi_status
acpi_execute_reg_methods(acpi_handle device, acpi_adr_space_type space_id)
acpi_execute_reg_methods(acpi_handle device, u32 max_depth,
acpi_adr_space_type space_id)
{
struct acpi_namespace_node *node;
acpi_status status;
@@ -296,7 +299,8 @@ acpi_execute_reg_methods(acpi_handle device, acpi_adr_space_type space_id)
/* Run all _REG methods for this address space */
acpi_ev_execute_reg_methods(node, space_id, ACPI_REG_CONNECT);
acpi_ev_execute_reg_methods(node, max_depth, space_id,
ACPI_REG_CONNECT);
} else {
status = AE_BAD_PARAMETER;
}
@@ -306,57 +310,3 @@ acpi_execute_reg_methods(acpi_handle device, acpi_adr_space_type space_id)
}
ACPI_EXPORT_SYMBOL(acpi_execute_reg_methods)
/*******************************************************************************
*
* FUNCTION: acpi_execute_orphan_reg_method
*
* PARAMETERS: device - Handle for the device
* space_id - The address space ID
*
* RETURN: Status
*
* DESCRIPTION: Execute an "orphan" _REG method that appears under an ACPI
* device. This is a _REG method that has no corresponding region
* within the device's scope.
*
******************************************************************************/
acpi_status
acpi_execute_orphan_reg_method(acpi_handle device, acpi_adr_space_type space_id)
{
struct acpi_namespace_node *node;
acpi_status status;
ACPI_FUNCTION_TRACE(acpi_execute_orphan_reg_method);
/* Parameter validation */
if (!device) {
return_ACPI_STATUS(AE_BAD_PARAMETER);
}
status = acpi_ut_acquire_mutex(ACPI_MTX_NAMESPACE);
if (ACPI_FAILURE(status)) {
return_ACPI_STATUS(status);
}
/* Convert and validate the device handle */
node = acpi_ns_validate_handle(device);
if (node) {
/*
* If an "orphan" _REG method is present in the device's scope
* for the given address space ID, run it.
*/
acpi_ev_execute_orphan_reg_method(node, space_id);
} else {
status = AE_BAD_PARAMETER;
}
(void)acpi_ut_release_mutex(ACPI_MTX_NAMESPACE);
return_ACPI_STATUS(status);
}
ACPI_EXPORT_SYMBOL(acpi_execute_orphan_reg_method)
drivers/acpi/ec.c
+9 -5
View File
@@ -1487,12 +1487,13 @@ static bool install_gpio_irq_event_handler(struct acpi_ec *ec)
static int ec_install_handlers(struct acpi_ec *ec, struct acpi_device *device,
bool call_reg)
{
acpi_handle scope_handle = ec == first_ec ? ACPI_ROOT_OBJECT : ec->handle;
acpi_status status;
acpi_ec_start(ec, false);
if (!test_bit(EC_FLAGS_EC_HANDLER_INSTALLED, &ec->flags)) {
acpi_handle scope_handle = ec == first_ec ? ACPI_ROOT_OBJECT : ec->handle;
acpi_ec_enter_noirq(ec);
status = acpi_install_address_space_handler_no_reg(scope_handle,
ACPI_ADR_SPACE_EC,
@@ -1506,10 +1507,7 @@ static int ec_install_handlers(struct acpi_ec *ec, struct acpi_device *device,
}
if (call_reg && !test_bit(EC_FLAGS_EC_REG_CALLED, &ec->flags)) {
acpi_execute_reg_methods(scope_handle, ACPI_ADR_SPACE_EC);
if (scope_handle != ec->handle)
acpi_execute_orphan_reg_method(ec->handle, ACPI_ADR_SPACE_EC);
acpi_execute_reg_methods(ec->handle, ACPI_UINT32_MAX, ACPI_ADR_SPACE_EC);
set_bit(EC_FLAGS_EC_REG_CALLED, &ec->flags);
}
@@ -1724,6 +1722,12 @@ static void acpi_ec_remove(struct acpi_device *device)
}
}
void acpi_ec_register_opregions(struct acpi_device *adev)
{
if (first_ec && first_ec->handle != adev->handle)
acpi_execute_reg_methods(adev->handle, 1, ACPI_ADR_SPACE_EC);
}
static acpi_status
ec_parse_io_ports(struct acpi_resource *resource, void *context)
{
drivers/acpi/internal.h
+1
View File
@@ -223,6 +223,7 @@ int acpi_ec_add_query_handler(struct acpi_ec *ec, u8 query_bit,
acpi_handle handle, acpi_ec_query_func func,
void *data);
void acpi_ec_remove_query_handler(struct acpi_ec *ec, u8 query_bit);
void acpi_ec_register_opregions(struct acpi_device *adev);
#ifdef CONFIG_PM_SLEEP
void acpi_ec_flush_work(void);
drivers/acpi/scan.c
+2
View File
@@ -2273,6 +2273,8 @@ static int acpi_bus_attach(struct acpi_device *device, void *first_pass)
if (device->handler)
goto ok;
acpi_ec_register_opregions(device);
if (!device->flags.initialized) {
device->flags.power_manageable =
device->power.states[ACPI_STATE_D0].flags.valid;
drivers/ata/libata-scsi.c
+13 -2
View File
@@ -951,8 +951,19 @@ static void ata_gen_passthru_sense(struct ata_queued_cmd *qc)
&sense_key, &asc, &ascq);
ata_scsi_set_sense(qc->dev, cmd, sense_key, asc, ascq);
} else {
/* ATA PASS-THROUGH INFORMATION AVAILABLE */
ata_scsi_set_sense(qc->dev, cmd, RECOVERED_ERROR, 0, 0x1D);
/*
* ATA PASS-THROUGH INFORMATION AVAILABLE
*
* Note: we are supposed to call ata_scsi_set_sense(), which
* respects the D_SENSE bit, instead of unconditionally
* generating the sense data in descriptor format. However,
* because hdparm, hddtemp, and udisks incorrectly assume sense
* data in descriptor format, without even looking at the
* RESPONSE CODE field in the returned sense data (to see which
* format the returned sense data is in), we are stuck with
* being bug compatible with older kernels.
*/
scsi_build_sense(cmd, 1, RECOVERED_ERROR, 0, 0x1D);
}
}
drivers/atm/idt77252.c
+5 -4
View File
@@ -1118,8 +1118,8 @@ dequeue_rx(struct idt77252_dev *card, struct rsq_entry *rsqe)
rpp->len += skb->len;
if (stat & SAR_RSQE_EPDU) {
unsigned int len, truesize;
unsigned char *l1l2;
unsigned int len;
l1l2 = (unsigned char *) ((unsigned long) skb->data + skb->len - 6);
@@ -1189,14 +1189,15 @@ dequeue_rx(struct idt77252_dev *card, struct rsq_entry *rsqe)
ATM_SKB(skb)->vcc = vcc;
__net_timestamp(skb);
truesize = skb->truesize;
vcc->push(vcc, skb);
atomic_inc(&vcc->stats->rx);
if (skb->truesize > SAR_FB_SIZE_3)
if (truesize > SAR_FB_SIZE_3)
add_rx_skb(card, 3, SAR_FB_SIZE_3, 1);
else if (skb->truesize > SAR_FB_SIZE_2)
else if (truesize > SAR_FB_SIZE_2)
add_rx_skb(card, 2, SAR_FB_SIZE_2, 1);
else if (skb->truesize > SAR_FB_SIZE_1)
else if (truesize > SAR_FB_SIZE_1)
add_rx_skb(card, 1, SAR_FB_SIZE_1, 1);
else
add_rx_skb(card, 0, SAR_FB_SIZE_0, 1);
drivers/char/xillybus/xillyusb.c
+34 -8
View File
@@ -50,6 +50,7 @@ MODULE_LICENSE("GPL v2");
static const char xillyname[] = "xillyusb";
static unsigned int fifo_buf_order;
static struct workqueue_struct *wakeup_wq;
#define USB_VENDOR_ID_XILINX 0x03fd
#define USB_VENDOR_ID_ALTERA 0x09fb
@@ -569,10 +570,6 @@ static void cleanup_dev(struct kref *kref)
* errors if executed. The mechanism relies on that xdev->error is assigned
* a non-zero value by report_io_error() prior to queueing wakeup_all(),
* which prevents bulk_in_work() from calling process_bulk_in().
*
* The fact that wakeup_all() and bulk_in_work() are queued on the same
* workqueue makes their concurrent execution very unlikely, however the
* kernel's API doesn't seem to ensure this strictly.
*/
static void wakeup_all(struct work_struct *work)
@@ -627,7 +624,7 @@ static void report_io_error(struct xillyusb_dev *xdev,
if (do_once) {
kref_get(&xdev->kref); /* xdev is used by work item */
queue_work(xdev->workq, &xdev->wakeup_workitem);
queue_work(wakeup_wq, &xdev->wakeup_workitem);
}
}
@@ -1906,6 +1903,13 @@ static const struct file_operations xillyusb_fops = {
static int xillyusb_setup_base_eps(struct xillyusb_dev *xdev)
{
struct usb_device *udev = xdev->udev;
/* Verify that device has the two fundamental bulk in/out endpoints */
if (usb_pipe_type_check(udev, usb_sndbulkpipe(udev, MSG_EP_NUM)) ||
usb_pipe_type_check(udev, usb_rcvbulkpipe(udev, IN_EP_NUM)))
return -ENODEV;
xdev->msg_ep = endpoint_alloc(xdev, MSG_EP_NUM | USB_DIR_OUT,
bulk_out_work, 1, 2);
if (!xdev->msg_ep)
@@ -1935,14 +1939,15 @@ static int setup_channels(struct xillyusb_dev *xdev,
__le16 *chandesc,
int num_channels)
{
struct xillyusb_channel *chan;
struct usb_device *udev = xdev->udev;
struct xillyusb_channel *chan, *new_channels;
int i;
chan = kcalloc(num_channels, sizeof(*chan), GFP_KERNEL);
if (!chan)
return -ENOMEM;
xdev->channels = chan;
new_channels = chan;
for (i = 0; i < num_channels; i++, chan++) {
unsigned int in_desc = le16_to_cpu(*chandesc++);
@@ -1971,6 +1976,15 @@ static int setup_channels(struct xillyusb_dev *xdev,
*/
if ((out_desc & 0x80) && i < 14) { /* Entry is valid */
if (usb_pipe_type_check(udev,
usb_sndbulkpipe(udev, i + 2))) {
dev_err(xdev->dev,
"Missing BULK OUT endpoint %d\n",
i + 2);
kfree(new_channels);
return -ENODEV;
}
chan->writable = 1;
chan->out_synchronous = !!(out_desc & 0x40);
chan->out_seekable = !!(out_desc & 0x20);
@@ -1980,6 +1994,7 @@ static int setup_channels(struct xillyusb_dev *xdev,
}
}
xdev->channels = new_channels;
return 0;
}
@@ -2096,9 +2111,11 @@ static int xillyusb_discovery(struct usb_interface *interface)
* just after responding with the IDT, there is no reason for any
* work item to be running now. To be sure that xdev->channels
* is updated on anything that might run in parallel, flush the
* workqueue, which rarely does anything.
* device's workqueue and the wakeup work item. This rarely
* does anything.
*/
flush_workqueue(xdev->workq);
flush_work(&xdev->wakeup_workitem);
xdev->num_channels = num_channels;
@@ -2258,6 +2275,10 @@ static int __init xillyusb_init(void)
{
int rc = 0;
wakeup_wq = alloc_workqueue(xillyname, 0, 0);
if (!wakeup_wq)
return -ENOMEM;
if (LOG2_INITIAL_FIFO_BUF_SIZE > PAGE_SHIFT)
fifo_buf_order = LOG2_INITIAL_FIFO_BUF_SIZE - PAGE_SHIFT;
else
@@ -2265,12 +2286,17 @@ static int __init xillyusb_init(void)
rc = usb_register(&xillyusb_driver);
if (rc)
destroy_workqueue(wakeup_wq);
return rc;
}
static void __exit xillyusb_exit(void)
{
usb_deregister(&xillyusb_driver);
destroy_workqueue(wakeup_wq);
}
module_init(xillyusb_init);
drivers/clk/thead/clk-th1520-ap.c
+1 -1
View File
@@ -738,7 +738,7 @@ static struct ccu_div vp_axi_clk = {
.hw.init = CLK_HW_INIT_PARENTS_HW("vp-axi",
video_pll_clk_parent,
&ccu_div_ops,
0),
CLK_IGNORE_UNUSED),
},
};
drivers/gpio/gpio-mlxbf3.c
+14
View File
@@ -39,6 +39,8 @@
#define MLXBF_GPIO_CAUSE_OR_EVTEN0 0x14
#define MLXBF_GPIO_CAUSE_OR_CLRCAUSE 0x18
#define MLXBF_GPIO_CLR_ALL_INTS GENMASK(31, 0)
struct mlxbf3_gpio_context {
struct gpio_chip gc;
@@ -82,6 +84,8 @@ static void mlxbf3_gpio_irq_disable(struct irq_data *irqd)
val = readl(gs->gpio_cause_io + MLXBF_GPIO_CAUSE_OR_EVTEN0);
val &= ~BIT(offset);
writel(val, gs->gpio_cause_io + MLXBF_GPIO_CAUSE_OR_EVTEN0);
writel(BIT(offset), gs->gpio_cause_io + MLXBF_GPIO_CAUSE_OR_CLRCAUSE);
raw_spin_unlock_irqrestore(&gs->gc.bgpio_lock, flags);
gpiochip_disable_irq(gc, offset);
@@ -253,6 +257,15 @@ static int mlxbf3_gpio_probe(struct platform_device *pdev)
return 0;
}
static void mlxbf3_gpio_shutdown(struct platform_device *pdev)
{
struct mlxbf3_gpio_context *gs = platform_get_drvdata(pdev);
/* Disable and clear all interrupts */
writel(0, gs->gpio_cause_io + MLXBF_GPIO_CAUSE_OR_EVTEN0);
writel(MLXBF_GPIO_CLR_ALL_INTS, gs->gpio_cause_io + MLXBF_GPIO_CAUSE_OR_CLRCAUSE);
}
static const struct acpi_device_id mlxbf3_gpio_acpi_match[] = {
{ "MLNXBF33", 0 },
{}
@@ -265,6 +278,7 @@ static struct platform_driver mlxbf3_gpio_driver = {
.acpi_match_table = mlxbf3_gpio_acpi_match,
},
.probe = mlxbf3_gpio_probe,
.shutdown = mlxbf3_gpio_shutdown,
};
module_platform_driver(mlxbf3_gpio_driver);
drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
+3
View File
@@ -1057,6 +1057,9 @@ static int amdgpu_cs_patch_ibs(struct amdgpu_cs_parser *p,
r = amdgpu_ring_parse_cs(ring, p, job, ib);
if (r)
return r;
if (ib->sa_bo)
ib->gpu_addr = amdgpu_sa_bo_gpu_addr(ib->sa_bo);
} else {
ib->ptr = (uint32_t *)kptr;
r = amdgpu_ring_patch_cs_in_place(ring, p, job, ib);
drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
+8
View File
@@ -685,16 +685,24 @@ int amdgpu_ctx_ioctl(struct drm_device *dev, void *data,
switch (args->in.op) {
case AMDGPU_CTX_OP_ALLOC_CTX:
if (args->in.flags)
return -EINVAL;
r = amdgpu_ctx_alloc(adev, fpriv, filp, priority, &id);
args->out.alloc.ctx_id = id;
break;
case AMDGPU_CTX_OP_FREE_CTX:
if (args->in.flags)
return -EINVAL;
r = amdgpu_ctx_free(fpriv, id);
break;
case AMDGPU_CTX_OP_QUERY_STATE:
if (args->in.flags)
return -EINVAL;
r = amdgpu_ctx_query(adev, fpriv, id, &args->out);
break;
case AMDGPU_CTX_OP_QUERY_STATE2:
if (args->in.flags)
return -EINVAL;
r = amdgpu_ctx_query2(adev, fpriv, id, &args->out);
break;
case AMDGPU_CTX_OP_GET_STABLE_PSTATE:
drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c
+24 -2
View File
@@ -509,6 +509,16 @@ int amdgpu_gfx_disable_kcq(struct amdgpu_device *adev, int xcc_id)
int i, r = 0;
int j;
if (adev->enable_mes) {
for (i = 0; i < adev->gfx.num_compute_rings; i++) {
j = i + xcc_id * adev->gfx.num_compute_rings;
amdgpu_mes_unmap_legacy_queue(adev,
&adev->gfx.compute_ring[j],
RESET_QUEUES, 0, 0);
}
return 0;
}
if (!kiq->pmf || !kiq->pmf->kiq_unmap_queues)
return -EINVAL;
@@ -551,6 +561,18 @@ int amdgpu_gfx_disable_kgq(struct amdgpu_device *adev, int xcc_id)
int i, r = 0;
int j;
if (adev->enable_mes) {
if (amdgpu_gfx_is_master_xcc(adev, xcc_id)) {
for (i = 0; i < adev->gfx.num_gfx_rings; i++) {
j = i + xcc_id * adev->gfx.num_gfx_rings;
amdgpu_mes_unmap_legacy_queue(adev,
&adev->gfx.gfx_ring[j],
PREEMPT_QUEUES, 0, 0);
}
}
return 0;
}
if (!kiq->pmf || !kiq->pmf->kiq_unmap_queues)
return -EINVAL;
@@ -995,7 +1017,7 @@ uint32_t amdgpu_kiq_rreg(struct amdgpu_device *adev, uint32_t reg, uint32_t xcc_
if (amdgpu_device_skip_hw_access(adev))
return 0;
if (adev->mes.ring.sched.ready)
if (adev->mes.ring[0].sched.ready)
return amdgpu_mes_rreg(adev, reg);
BUG_ON(!ring->funcs->emit_rreg);
@@ -1065,7 +1087,7 @@ void amdgpu_kiq_wreg(struct amdgpu_device *adev, uint32_t reg, uint32_t v, uint3
if (amdgpu_device_skip_hw_access(adev))
return;
if (adev->mes.ring.sched.ready) {
if (adev->mes.ring[0].sched.ready) {
amdgpu_mes_wreg(adev, reg, v);
return;
}
drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c
+3 -2
View File
@@ -589,7 +589,8 @@ int amdgpu_gmc_allocate_vm_inv_eng(struct amdgpu_device *adev)
ring = adev->rings[i];
vmhub = ring->vm_hub;
if (ring == &adev->mes.ring ||
if (ring == &adev->mes.ring[0] ||
ring == &adev->mes.ring[1] ||
ring == &adev->umsch_mm.ring)
continue;
@@ -761,7 +762,7 @@ void amdgpu_gmc_fw_reg_write_reg_wait(struct amdgpu_device *adev,
unsigned long flags;
uint32_t seq;
if (adev->mes.ring.sched.ready) {
if (adev->mes.ring[0].sched.ready) {
amdgpu_mes_reg_write_reg_wait(adev, reg0, reg1,
ref, mask);
return;
drivers/gpu/drm/amd/amdgpu/amdgpu_mes.c
+51 -32
View File
@@ -135,9 +135,11 @@ int amdgpu_mes_init(struct amdgpu_device *adev)
idr_init(&adev->mes.queue_id_idr);
ida_init(&adev->mes.doorbell_ida);
spin_lock_init(&adev->mes.queue_id_lock);
spin_lock_init(&adev->mes.ring_lock);
mutex_init(&adev->mes.mutex_hidden);
for (i = 0; i < AMDGPU_MAX_MES_PIPES; i++)
spin_lock_init(&adev->mes.ring_lock[i]);
adev->mes.total_max_queue = AMDGPU_FENCE_MES_QUEUE_ID_MASK;
adev->mes.vmid_mask_mmhub = 0xffffff00;
adev->mes.vmid_mask_gfxhub = 0xffffff00;
@@ -163,36 +165,38 @@ int amdgpu_mes_init(struct amdgpu_device *adev)
adev->mes.sdma_hqd_mask[i] = 0xfc;
}
r = amdgpu_device_wb_get(adev, &adev->mes.sch_ctx_offs);
if (r) {
dev_err(adev->dev,
"(%d) ring trail_fence_offs wb alloc failed\n", r);
goto error_ids;
}
adev->mes.sch_ctx_gpu_addr =
adev->wb.gpu_addr + (adev->mes.sch_ctx_offs * 4);
adev->mes.sch_ctx_ptr =
(uint64_t *)&adev->wb.wb[adev->mes.sch_ctx_offs];
for (i = 0; i < AMDGPU_MAX_MES_PIPES; i++) {
r = amdgpu_device_wb_get(adev, &adev->mes.sch_ctx_offs[i]);
if (r) {
dev_err(adev->dev,
"(%d) ring trail_fence_offs wb alloc failed\n",
r);
goto error;
}
adev->mes.sch_ctx_gpu_addr[i] =
adev->wb.gpu_addr + (adev->mes.sch_ctx_offs[i] * 4);
adev->mes.sch_ctx_ptr[i] =
(uint64_t *)&adev->wb.wb[adev->mes.sch_ctx_offs[i]];
r = amdgpu_device_wb_get(adev, &adev->mes.query_status_fence_offs);
if (r) {
amdgpu_device_wb_free(adev, adev->mes.sch_ctx_offs);
dev_err(adev->dev,
"(%d) query_status_fence_offs wb alloc failed\n", r);
goto error_ids;
r = amdgpu_device_wb_get(adev,
&adev->mes.query_status_fence_offs[i]);
if (r) {
dev_err(adev->dev,
"(%d) query_status_fence_offs wb alloc failed\n",
r);
goto error;
}
adev->mes.query_status_fence_gpu_addr[i] = adev->wb.gpu_addr +
(adev->mes.query_status_fence_offs[i] * 4);
adev->mes.query_status_fence_ptr[i] =
(uint64_t *)&adev->wb.wb[adev->mes.query_status_fence_offs[i]];
}
adev->mes.query_status_fence_gpu_addr =
adev->wb.gpu_addr + (adev->mes.query_status_fence_offs * 4);
adev->mes.query_status_fence_ptr =
(uint64_t *)&adev->wb.wb[adev->mes.query_status_fence_offs];
r = amdgpu_device_wb_get(adev, &adev->mes.read_val_offs);
if (r) {
amdgpu_device_wb_free(adev, adev->mes.sch_ctx_offs);
amdgpu_device_wb_free(adev, adev->mes.query_status_fence_offs);
dev_err(adev->dev,
"(%d) read_val_offs alloc failed\n", r);
goto error_ids;
goto error;
}
adev->mes.read_val_gpu_addr =
adev->wb.gpu_addr + (adev->mes.read_val_offs * 4);
@@ -212,10 +216,16 @@ int amdgpu_mes_init(struct amdgpu_device *adev)
error_doorbell:
amdgpu_mes_doorbell_free(adev);
error:
amdgpu_device_wb_free(adev, adev->mes.sch_ctx_offs);
amdgpu_device_wb_free(adev, adev->mes.query_status_fence_offs);
amdgpu_device_wb_free(adev, adev->mes.read_val_offs);
error_ids:
for (i = 0; i < AMDGPU_MAX_MES_PIPES; i++) {
if (adev->mes.sch_ctx_ptr[i])
amdgpu_device_wb_free(adev, adev->mes.sch_ctx_offs[i]);
if (adev->mes.query_status_fence_ptr[i])
amdgpu_device_wb_free(adev,
adev->mes.query_status_fence_offs[i]);
}
if (adev->mes.read_val_ptr)
amdgpu_device_wb_free(adev, adev->mes.read_val_offs);
idr_destroy(&adev->mes.pasid_idr);
idr_destroy(&adev->mes.gang_id_idr);
idr_destroy(&adev->mes.queue_id_idr);
@@ -226,13 +236,22 @@ error_ids:
void amdgpu_mes_fini(struct amdgpu_device *adev)
{
int i;
amdgpu_bo_free_kernel(&adev->mes.event_log_gpu_obj,
&adev->mes.event_log_gpu_addr,
&adev->mes.event_log_cpu_addr);
amdgpu_device_wb_free(adev, adev->mes.sch_ctx_offs);
amdgpu_device_wb_free(adev, adev->mes.query_status_fence_offs);
amdgpu_device_wb_free(adev, adev->mes.read_val_offs);
for (i = 0; i < AMDGPU_MAX_MES_PIPES; i++) {
if (adev->mes.sch_ctx_ptr[i])
amdgpu_device_wb_free(adev, adev->mes.sch_ctx_offs[i]);
if (adev->mes.query_status_fence_ptr[i])
amdgpu_device_wb_free(adev,
adev->mes.query_status_fence_offs[i]);
}
if (adev->mes.read_val_ptr)
amdgpu_device_wb_free(adev, adev->mes.read_val_offs);
amdgpu_mes_doorbell_free(adev);
idr_destroy(&adev->mes.pasid_idr);
@@ -1499,7 +1518,7 @@ int amdgpu_mes_init_microcode(struct amdgpu_device *adev, int pipe)
amdgpu_ucode_ip_version_decode(adev, GC_HWIP, ucode_prefix,
sizeof(ucode_prefix));
if (adev->enable_uni_mes && pipe == AMDGPU_MES_SCHED_PIPE) {
if (adev->enable_uni_mes) {
snprintf(fw_name, sizeof(fw_name),
"amdgpu/%s_uni_mes.bin", ucode_prefix);
} else if (amdgpu_ip_version(adev, GC_HWIP, 0) >= IP_VERSION(11, 0, 0) &&
drivers/gpu/drm/amd/amdgpu/amdgpu_mes.h
+8 -8
View File
@@ -82,8 +82,8 @@ struct amdgpu_mes {
uint64_t default_process_quantum;
uint64_t default_gang_quantum;
struct amdgpu_ring ring;
spinlock_t ring_lock;
struct amdgpu_ring ring[AMDGPU_MAX_MES_PIPES];
spinlock_t ring_lock[AMDGPU_MAX_MES_PIPES];
const struct firmware *fw[AMDGPU_MAX_MES_PIPES];
@@ -112,12 +112,12 @@ struct amdgpu_mes {
uint32_t gfx_hqd_mask[AMDGPU_MES_MAX_GFX_PIPES];
uint32_t sdma_hqd_mask[AMDGPU_MES_MAX_SDMA_PIPES];
uint32_t aggregated_doorbells[AMDGPU_MES_PRIORITY_NUM_LEVELS];
uint32_t sch_ctx_offs;
uint64_t sch_ctx_gpu_addr;
uint64_t *sch_ctx_ptr;
uint32_t query_status_fence_offs;
uint64_t query_status_fence_gpu_addr;
uint64_t *query_status_fence_ptr;
uint32_t sch_ctx_offs[AMDGPU_MAX_MES_PIPES];
uint64_t sch_ctx_gpu_addr[AMDGPU_MAX_MES_PIPES];
uint64_t *sch_ctx_ptr[AMDGPU_MAX_MES_PIPES];
uint32_t query_status_fence_offs[AMDGPU_MAX_MES_PIPES];
uint64_t query_status_fence_gpu_addr[AMDGPU_MAX_MES_PIPES];
uint64_t *query_status_fence_ptr[AMDGPU_MAX_MES_PIPES];
uint32_t read_val_offs;
uint64_t read_val_gpu_addr;
uint32_t *read_val_ptr;
drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c
+2
View File
@@ -212,6 +212,8 @@ int amdgpu_ring_init(struct amdgpu_device *adev, struct amdgpu_ring *ring,
*/
if (ring->funcs->type == AMDGPU_RING_TYPE_KIQ)
sched_hw_submission = max(sched_hw_submission, 256);
if (ring->funcs->type == AMDGPU_RING_TYPE_MES)
sched_hw_submission = 8;
else if (ring == &adev->sdma.instance[0].page)
sched_hw_submission = 256;
drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h
+4 -1
View File
@@ -461,8 +461,11 @@ struct amdgpu_vcn5_fw_shared {
struct amdgpu_fw_shared_unified_queue_struct sq;
uint8_t pad1[8];
struct amdgpu_fw_shared_fw_logging fw_log;
uint8_t pad2[20];
struct amdgpu_fw_shared_rb_setup rb_setup;
uint8_t pad2[4];
struct amdgpu_fw_shared_smu_interface_info smu_dpm_interface;
struct amdgpu_fw_shared_drm_key_wa drm_key_wa;
uint8_t pad3[9];
};
#define VCN_BLOCK_ENCODE_DISABLE_MASK 0x80
drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c
+1 -1
View File
@@ -858,7 +858,7 @@ void amdgpu_virt_post_reset(struct amdgpu_device *adev)
adev->gfx.is_poweron = false;
}
adev->mes.ring.sched.ready = false;
adev->mes.ring[0].sched.ready = false;
}
bool amdgpu_virt_fw_load_skip_check(struct amdgpu_device *adev, uint32_t ucode_id)
drivers/gpu/drm/amd/amdgpu/gfx_v12_0.c
+1 -26
View File
@@ -3546,33 +3546,9 @@ static int gfx_v12_0_hw_init(void *handle)
return r;
}
static int gfx_v12_0_kiq_disable_kgq(struct amdgpu_device *adev)
{
struct amdgpu_kiq *kiq = &adev->gfx.kiq[0];
struct amdgpu_ring *kiq_ring = &kiq->ring;
int i, r = 0;
if (!kiq->pmf || !kiq->pmf->kiq_unmap_queues)
return -EINVAL;
if (amdgpu_ring_alloc(kiq_ring, kiq->pmf->unmap_queues_size *
adev->gfx.num_gfx_rings))
return -ENOMEM;
for (i = 0; i < adev->gfx.num_gfx_rings; i++)
kiq->pmf->kiq_unmap_queues(kiq_ring, &adev->gfx.gfx_ring[i],
PREEMPT_QUEUES, 0, 0);
if (adev->gfx.kiq[0].ring.sched.ready)
r = amdgpu_ring_test_helper(kiq_ring);
return r;
}
static int gfx_v12_0_hw_fini(void *handle)
{
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
int r;
uint32_t tmp;
amdgpu_irq_put(adev, &adev->gfx.priv_reg_irq, 0);
@@ -3580,8 +3556,7 @@ static int gfx_v12_0_hw_fini(void *handle)
if (!adev->no_hw_access) {
if (amdgpu_async_gfx_ring) {
r = gfx_v12_0_kiq_disable_kgq(adev);
if (r)
if (amdgpu_gfx_disable_kgq(adev, 0))
DRM_ERROR("KGQ disable failed\n");
}
drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c
+1 -1
View File
@@ -231,7 +231,7 @@ static void gmc_v11_0_flush_gpu_tlb(struct amdgpu_device *adev, uint32_t vmid,
/* This is necessary for SRIOV as well as for GFXOFF to function
* properly under bare metal
*/
if ((adev->gfx.kiq[0].ring.sched.ready || adev->mes.ring.sched.ready) &&
if ((adev->gfx.kiq[0].ring.sched.ready || adev->mes.ring[0].sched.ready) &&
(amdgpu_sriov_runtime(adev) || !amdgpu_sriov_vf(adev))) {
amdgpu_gmc_fw_reg_write_reg_wait(adev, req, ack, inv_req,
1 << vmid, GET_INST(GC, 0));
drivers/gpu/drm/amd/amdgpu/gmc_v12_0.c
+1 -1
View File
@@ -299,7 +299,7 @@ static void gmc_v12_0_flush_gpu_tlb(struct amdgpu_device *adev, uint32_t vmid,
/* This is necessary for SRIOV as well as for GFXOFF to function
* properly under bare metal
*/
if ((adev->gfx.kiq[0].ring.sched.ready || adev->mes.ring.sched.ready) &&
if ((adev->gfx.kiq[0].ring.sched.ready || adev->mes.ring[0].sched.ready) &&
(amdgpu_sriov_runtime(adev) || !amdgpu_sriov_vf(adev))) {
struct amdgpu_vmhub *hub = &adev->vmhub[vmhub];
const unsigned eng = 17;
drivers/gpu/drm/amd/amdgpu/jpeg_v2_0.c
+2 -2
View File
@@ -538,11 +538,11 @@ void jpeg_v2_0_dec_ring_emit_ib(struct amdgpu_ring *ring,
amdgpu_ring_write(ring, PACKETJ(mmUVD_LMI_JRBC_IB_VMID_INTERNAL_OFFSET,
0, 0, PACKETJ_TYPE0));
amdgpu_ring_write(ring, (vmid | (vmid << 4)));
amdgpu_ring_write(ring, (vmid | (vmid << 4) | (vmid << 8)));
amdgpu_ring_write(ring, PACKETJ(mmUVD_LMI_JPEG_VMID_INTERNAL_OFFSET,
0, 0, PACKETJ_TYPE0));
amdgpu_ring_write(ring, (vmid | (vmid << 4)));
amdgpu_ring_write(ring, (vmid | (vmid << 4) | (vmid << 8)));
amdgpu_ring_write(ring, PACKETJ(mmUVD_LMI_JRBC_IB_64BIT_BAR_LOW_INTERNAL_OFFSET,
0, 0, PACKETJ_TYPE0));
drivers/gpu/drm/amd/amdgpu/jpeg_v4_0_3.c
+61 -2
View File
@@ -23,6 +23,7 @@
#include "amdgpu.h"
#include "amdgpu_jpeg.h"
#include "amdgpu_cs.h"
#include "soc15.h"
#include "soc15d.h"
#include "jpeg_v4_0_3.h"
@@ -782,11 +783,15 @@ void jpeg_v4_0_3_dec_ring_emit_ib(struct amdgpu_ring *ring,
amdgpu_ring_write(ring, PACKETJ(regUVD_LMI_JRBC_IB_VMID_INTERNAL_OFFSET,
0, 0, PACKETJ_TYPE0));
amdgpu_ring_write(ring, (vmid | (vmid << 4)));
if (ring->funcs->parse_cs)
amdgpu_ring_write(ring, 0);
else
amdgpu_ring_write(ring, (vmid | (vmid << 4) | (vmid << 8)));
amdgpu_ring_write(ring, PACKETJ(regUVD_LMI_JPEG_VMID_INTERNAL_OFFSET,
0, 0, PACKETJ_TYPE0));
amdgpu_ring_write(ring, (vmid | (vmid << 4)));
amdgpu_ring_write(ring, (vmid | (vmid << 4) | (vmid << 8)));
amdgpu_ring_write(ring, PACKETJ(regUVD_LMI_JRBC_IB_64BIT_BAR_LOW_INTERNAL_OFFSET,
0, 0, PACKETJ_TYPE0));
@@ -1084,6 +1089,7 @@ static const struct amdgpu_ring_funcs jpeg_v4_0_3_dec_ring_vm_funcs = {
.get_rptr = jpeg_v4_0_3_dec_ring_get_rptr,
.get_wptr = jpeg_v4_0_3_dec_ring_get_wptr,
.set_wptr = jpeg_v4_0_3_dec_ring_set_wptr,
.parse_cs = jpeg_v4_0_3_dec_ring_parse_cs,
.emit_frame_size =
SOC15_FLUSH_GPU_TLB_NUM_WREG * 6 +
SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 8 +
@@ -1248,3 +1254,56 @@ static void jpeg_v4_0_3_set_ras_funcs(struct amdgpu_device *adev)
{
adev->jpeg.ras = &jpeg_v4_0_3_ras;
}
/**
* jpeg_v4_0_3_dec_ring_parse_cs - command submission parser
*
* @parser: Command submission parser context
* @job: the job to parse
* @ib: the IB to parse
*
* Parse the command stream, return -EINVAL for invalid packet,
* 0 otherwise
*/
int jpeg_v4_0_3_dec_ring_parse_cs(struct amdgpu_cs_parser *parser,
struct amdgpu_job *job,
struct amdgpu_ib *ib)
{
uint32_t i, reg, res, cond, type;
struct amdgpu_device *adev = parser->adev;
for (i = 0; i < ib->length_dw ; i += 2) {
reg = CP_PACKETJ_GET_REG(ib->ptr[i]);
res = CP_PACKETJ_GET_RES(ib->ptr[i]);
cond = CP_PACKETJ_GET_COND(ib->ptr[i]);
type = CP_PACKETJ_GET_TYPE(ib->ptr[i]);
if (res) /* only support 0 at the moment */
return -EINVAL;
switch (type) {
case PACKETJ_TYPE0:
if (cond != PACKETJ_CONDITION_CHECK0 || reg < JPEG_REG_RANGE_START || reg > JPEG_REG_RANGE_END) {
dev_err(adev->dev, "Invalid packet [0x%08x]!\n", ib->ptr[i]);
return -EINVAL;
}
break;
case PACKETJ_TYPE3:
if (cond != PACKETJ_CONDITION_CHECK3 || reg < JPEG_REG_RANGE_START || reg > JPEG_REG_RANGE_END) {
dev_err(adev->dev, "Invalid packet [0x%08x]!\n", ib->ptr[i]);
return -EINVAL;
}
break;
case PACKETJ_TYPE6:
if (ib->ptr[i] == CP_PACKETJ_NOP)
continue;
dev_err(adev->dev, "Invalid packet [0x%08x]!\n", ib->ptr[i]);
return -EINVAL;
default:
dev_err(adev->dev, "Unknown packet type %d !\n", type);
return -EINVAL;
}
}
return 0;
}
drivers/gpu/drm/amd/amdgpu/jpeg_v4_0_3.h
+6 -1
View File
@@ -46,6 +46,9 @@
#define JRBC_DEC_EXTERNAL_REG_WRITE_ADDR 0x18000
#define JPEG_REG_RANGE_START 0x4000
#define JPEG_REG_RANGE_END 0x41c2
extern const struct amdgpu_ip_block_version jpeg_v4_0_3_ip_block;
void jpeg_v4_0_3_dec_ring_emit_ib(struct amdgpu_ring *ring,
@@ -62,5 +65,7 @@ void jpeg_v4_0_3_dec_ring_insert_end(struct amdgpu_ring *ring);
void jpeg_v4_0_3_dec_ring_emit_wreg(struct amdgpu_ring *ring, uint32_t reg, uint32_t val);
void jpeg_v4_0_3_dec_ring_emit_reg_wait(struct amdgpu_ring *ring, uint32_t reg,
uint32_t val, uint32_t mask);
int jpeg_v4_0_3_dec_ring_parse_cs(struct amdgpu_cs_parser *parser,
struct amdgpu_job *job,
struct amdgpu_ib *ib);
#endif /* __JPEG_V4_0_3_H__ */
drivers/gpu/drm/amd/amdgpu/jpeg_v5_0_0.c
+1
View File
@@ -646,6 +646,7 @@ static const struct amdgpu_ring_funcs jpeg_v5_0_0_dec_ring_vm_funcs = {
.get_rptr = jpeg_v5_0_0_dec_ring_get_rptr,
.get_wptr = jpeg_v5_0_0_dec_ring_get_wptr,
.set_wptr = jpeg_v5_0_0_dec_ring_set_wptr,
.parse_cs = jpeg_v4_0_3_dec_ring_parse_cs,
.emit_frame_size =
SOC15_FLUSH_GPU_TLB_NUM_WREG * 6 +
SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 8 +
drivers/gpu/drm/amd/amdgpu/mes_v11_0.c
+33 -26
View File
@@ -162,13 +162,13 @@ static int mes_v11_0_submit_pkt_and_poll_completion(struct amdgpu_mes *mes,
union MESAPI__QUERY_MES_STATUS mes_status_pkt;
signed long timeout = 3000000; /* 3000 ms */
struct amdgpu_device *adev = mes->adev;
struct amdgpu_ring *ring = &mes->ring;
struct amdgpu_ring *ring = &mes->ring[0];
struct MES_API_STATUS *api_status;
union MESAPI__MISC *x_pkt = pkt;
const char *op_str, *misc_op_str;
unsigned long flags;
u64 status_gpu_addr;
u32 status_offset;
u32 seq, status_offset;
u64 *status_ptr;
signed long r;
int ret;
@@ -191,11 +191,18 @@ static int mes_v11_0_submit_pkt_and_poll_completion(struct amdgpu_mes *mes,
status_ptr = (u64 *)&adev->wb.wb[status_offset];
*status_ptr = 0;
spin_lock_irqsave(&mes->ring_lock, flags);
spin_lock_irqsave(&mes->ring_lock[0], flags);
r = amdgpu_ring_alloc(ring, (size + sizeof(mes_status_pkt)) / 4);
if (r)
goto error_unlock_free;
seq = ++ring->fence_drv.sync_seq;
r = amdgpu_fence_wait_polling(ring,
seq - ring->fence_drv.num_fences_mask,
timeout);
if (r < 1)
goto error_undo;
api_status = (struct MES_API_STATUS *)((char *)pkt + api_status_off);
api_status->api_completion_fence_addr = status_gpu_addr;
api_status->api_completion_fence_value = 1;
@@ -208,14 +215,13 @@ static int mes_v11_0_submit_pkt_and_poll_completion(struct amdgpu_mes *mes,
mes_status_pkt.header.dwsize = API_FRAME_SIZE_IN_DWORDS;
mes_status_pkt.api_status.api_completion_fence_addr =
ring->fence_drv.gpu_addr;
mes_status_pkt.api_status.api_completion_fence_value =
++ring->fence_drv.sync_seq;
mes_status_pkt.api_status.api_completion_fence_value = seq;
amdgpu_ring_write_multiple(ring, &mes_status_pkt,
sizeof(mes_status_pkt) / 4);
amdgpu_ring_commit(ring);
spin_unlock_irqrestore(&mes->ring_lock, flags);
spin_unlock_irqrestore(&mes->ring_lock[0], flags);
op_str = mes_v11_0_get_op_string(x_pkt);
misc_op_str = mes_v11_0_get_misc_op_string(x_pkt);
@@ -229,7 +235,7 @@ static int mes_v11_0_submit_pkt_and_poll_completion(struct amdgpu_mes *mes,
dev_dbg(adev->dev, "MES msg=%d was emitted\n",
x_pkt->header.opcode);
r = amdgpu_fence_wait_polling(ring, ring->fence_drv.sync_seq, timeout);
r = amdgpu_fence_wait_polling(ring, seq, timeout);
if (r < 1 || !*status_ptr) {
if (misc_op_str)
@@ -252,8 +258,12 @@ static int mes_v11_0_submit_pkt_and_poll_completion(struct amdgpu_mes *mes,
amdgpu_device_wb_free(adev, status_offset);
return 0;
error_undo:
dev_err(adev->dev, "MES ring buffer is full.\n");
amdgpu_ring_undo(ring);
error_unlock_free:
spin_unlock_irqrestore(&mes->ring_lock, flags);
spin_unlock_irqrestore(&mes->ring_lock[0], flags);
error_wb_free:
amdgpu_device_wb_free(adev, status_offset);
@@ -512,9 +522,9 @@ static int mes_v11_0_set_hw_resources(struct amdgpu_mes *mes)
mes_set_hw_res_pkt.vmid_mask_gfxhub = mes->vmid_mask_gfxhub;
mes_set_hw_res_pkt.gds_size = adev->gds.gds_size;
mes_set_hw_res_pkt.paging_vmid = 0;
mes_set_hw_res_pkt.g_sch_ctx_gpu_mc_ptr = mes->sch_ctx_gpu_addr;
mes_set_hw_res_pkt.g_sch_ctx_gpu_mc_ptr = mes->sch_ctx_gpu_addr[0];
mes_set_hw_res_pkt.query_status_fence_gpu_mc_ptr =
mes->query_status_fence_gpu_addr;
mes->query_status_fence_gpu_addr[0];
for (i = 0; i < MAX_COMPUTE_PIPES; i++)
mes_set_hw_res_pkt.compute_hqd_mask[i] =
@@ -1015,7 +1025,7 @@ static int mes_v11_0_kiq_enable_queue(struct amdgpu_device *adev)
return r;
}
kiq->pmf->kiq_map_queues(kiq_ring, &adev->mes.ring);
kiq->pmf->kiq_map_queues(kiq_ring, &adev->mes.ring[0]);
return amdgpu_ring_test_helper(kiq_ring);
}
@@ -1029,7 +1039,7 @@ static int mes_v11_0_queue_init(struct amdgpu_device *adev,
if (pipe == AMDGPU_MES_KIQ_PIPE)
ring = &adev->gfx.kiq[0].ring;
else if (pipe == AMDGPU_MES_SCHED_PIPE)
ring = &adev->mes.ring;
ring = &adev->mes.ring[0];
else
BUG();
@@ -1071,7 +1081,7 @@ static int mes_v11_0_ring_init(struct amdgpu_device *adev)
{
struct amdgpu_ring *ring;
ring = &adev->mes.ring;
ring = &adev->mes.ring[0];
ring->funcs = &mes_v11_0_ring_funcs;
@@ -1124,7 +1134,7 @@ static int mes_v11_0_mqd_sw_init(struct amdgpu_device *adev,
if (pipe == AMDGPU_MES_KIQ_PIPE)
ring = &adev->gfx.kiq[0].ring;
else if (pipe == AMDGPU_MES_SCHED_PIPE)
ring = &adev->mes.ring;
ring = &adev->mes.ring[0];
else
BUG();
@@ -1200,9 +1210,6 @@ static int mes_v11_0_sw_fini(void *handle)
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
int pipe;
amdgpu_device_wb_free(adev, adev->mes.sch_ctx_offs);
amdgpu_device_wb_free(adev, adev->mes.query_status_fence_offs);
for (pipe = 0; pipe < AMDGPU_MAX_MES_PIPES; pipe++) {
kfree(adev->mes.mqd_backup[pipe]);
@@ -1216,12 +1223,12 @@ static int mes_v11_0_sw_fini(void *handle)
&adev->gfx.kiq[0].ring.mqd_gpu_addr,
&adev->gfx.kiq[0].ring.mqd_ptr);
amdgpu_bo_free_kernel(&adev->mes.ring.mqd_obj,
&adev->mes.ring.mqd_gpu_addr,
&adev->mes.ring.mqd_ptr);
amdgpu_bo_free_kernel(&adev->mes.ring[0].mqd_obj,
&adev->mes.ring[0].mqd_gpu_addr,
&adev->mes.ring[0].mqd_ptr);
amdgpu_ring_fini(&adev->gfx.kiq[0].ring);
amdgpu_ring_fini(&adev->mes.ring);
amdgpu_ring_fini(&adev->mes.ring[0]);
if (adev->firmware.load_type == AMDGPU_FW_LOAD_DIRECT) {
mes_v11_0_free_ucode_buffers(adev, AMDGPU_MES_KIQ_PIPE);
@@ -1332,9 +1339,9 @@ failure:
static int mes_v11_0_kiq_hw_fini(struct amdgpu_device *adev)
{
if (adev->mes.ring.sched.ready) {
mes_v11_0_kiq_dequeue(&adev->mes.ring);
adev->mes.ring.sched.ready = false;
if (adev->mes.ring[0].sched.ready) {
mes_v11_0_kiq_dequeue(&adev->mes.ring[0]);
adev->mes.ring[0].sched.ready = false;
}
if (amdgpu_sriov_vf(adev)) {
@@ -1352,7 +1359,7 @@ static int mes_v11_0_hw_init(void *handle)
int r;
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
if (adev->mes.ring.sched.ready)
if (adev->mes.ring[0].sched.ready)
goto out;
if (!adev->enable_mes_kiq) {
@@ -1397,7 +1404,7 @@ out:
* with MES enabled.
*/
adev->gfx.kiq[0].ring.sched.ready = false;
adev->mes.ring.sched.ready = true;
adev->mes.ring[0].sched.ready = true;
return 0;
drivers/gpu/drm/amd/amdgpu/mes_v12_0.c
+161 -135
View File
@@ -142,19 +142,20 @@ static const char *mes_v12_0_get_misc_op_string(union MESAPI__MISC *x_pkt)
}
static int mes_v12_0_submit_pkt_and_poll_completion(struct amdgpu_mes *mes,
void *pkt, int size,
int api_status_off)
int pipe, void *pkt, int size,
int api_status_off)
{
union MESAPI__QUERY_MES_STATUS mes_status_pkt;
signed long timeout = 3000000; /* 3000 ms */
struct amdgpu_device *adev = mes->adev;
struct amdgpu_ring *ring = &mes->ring;
struct amdgpu_ring *ring = &mes->ring[pipe];
spinlock_t *ring_lock = &mes->ring_lock[pipe];
struct MES_API_STATUS *api_status;
union MESAPI__MISC *x_pkt = pkt;
const char *op_str, *misc_op_str;
unsigned long flags;
u64 status_gpu_addr;
u32 status_offset;
u32 seq, status_offset;
u64 *status_ptr;
signed long r;
int ret;
@@ -177,11 +178,18 @@ static int mes_v12_0_submit_pkt_and_poll_completion(struct amdgpu_mes *mes,
status_ptr = (u64 *)&adev->wb.wb[status_offset];
*status_ptr = 0;
spin_lock_irqsave(&mes->ring_lock, flags);
spin_lock_irqsave(ring_lock, flags);
r = amdgpu_ring_alloc(ring, (size + sizeof(mes_status_pkt)) / 4);
if (r)
goto error_unlock_free;
seq = ++ring->fence_drv.sync_seq;
r = amdgpu_fence_wait_polling(ring,
seq - ring->fence_drv.num_fences_mask,
timeout);
if (r < 1)
goto error_undo;
api_status = (struct MES_API_STATUS *)((char *)pkt + api_status_off);
api_status->api_completion_fence_addr = status_gpu_addr;
api_status->api_completion_fence_value = 1;
@@ -194,39 +202,39 @@ static int mes_v12_0_submit_pkt_and_poll_completion(struct amdgpu_mes *mes,
mes_status_pkt.header.dwsize = API_FRAME_SIZE_IN_DWORDS;
mes_status_pkt.api_status.api_completion_fence_addr =
ring->fence_drv.gpu_addr;
mes_status_pkt.api_status.api_completion_fence_value =
++ring->fence_drv.sync_seq;
mes_status_pkt.api_status.api_completion_fence_value = seq;
amdgpu_ring_write_multiple(ring, &mes_status_pkt,
sizeof(mes_status_pkt) / 4);
amdgpu_ring_commit(ring);
spin_unlock_irqrestore(&mes->ring_lock, flags);
spin_unlock_irqrestore(ring_lock, flags);
op_str = mes_v12_0_get_op_string(x_pkt);
misc_op_str = mes_v12_0_get_misc_op_string(x_pkt);
if (misc_op_str)
dev_dbg(adev->dev, "MES msg=%s (%s) was emitted\n", op_str,
misc_op_str);
dev_dbg(adev->dev, "MES(%d) msg=%s (%s) was emitted\n",
pipe, op_str, misc_op_str);
else if (op_str)
dev_dbg(adev->dev, "MES msg=%s was emitted\n", op_str);
dev_dbg(adev->dev, "MES(%d) msg=%s was emitted\n",
pipe, op_str);
else
dev_dbg(adev->dev, "MES msg=%d was emitted\n",
x_pkt->header.opcode);
dev_dbg(adev->dev, "MES(%d) msg=%d was emitted\n",
pipe, x_pkt->header.opcode);
r = amdgpu_fence_wait_polling(ring, ring->fence_drv.sync_seq, timeout);
r = amdgpu_fence_wait_polling(ring, seq, timeout);
if (r < 1 || !*status_ptr) {
if (misc_op_str)
dev_err(adev->dev, "MES failed to respond to msg=%s (%s)\n",
op_str, misc_op_str);
dev_err(adev->dev, "MES(%d) failed to respond to msg=%s (%s)\n",
pipe, op_str, misc_op_str);
else if (op_str)
dev_err(adev->dev, "MES failed to respond to msg=%s\n",
op_str);
dev_err(adev->dev, "MES(%d) failed to respond to msg=%s\n",
pipe, op_str);
else
dev_err(adev->dev, "MES failed to respond to msg=%d\n",
x_pkt->header.opcode);
dev_err(adev->dev, "MES(%d) failed to respond to msg=%d\n",
pipe, x_pkt->header.opcode);
while (halt_if_hws_hang)
schedule();
@@ -238,8 +246,12 @@ static int mes_v12_0_submit_pkt_and_poll_completion(struct amdgpu_mes *mes,
amdgpu_device_wb_free(adev, status_offset);
return 0;
error_undo:
dev_err(adev->dev, "MES ring buffer is full.\n");
amdgpu_ring_undo(ring);
error_unlock_free:
spin_unlock_irqrestore(&mes->ring_lock, flags);
spin_unlock_irqrestore(ring_lock, flags);
error_wb_free:
amdgpu_device_wb_free(adev, status_offset);
@@ -254,6 +266,8 @@ static int convert_to_mes_queue_type(int queue_type)
return MES_QUEUE_TYPE_COMPUTE;
else if (queue_type == AMDGPU_RING_TYPE_SDMA)
return MES_QUEUE_TYPE_SDMA;
else if (queue_type == AMDGPU_RING_TYPE_MES)
return MES_QUEUE_TYPE_SCHQ;
else
BUG();
return -1;
@@ -311,6 +325,7 @@ static int mes_v12_0_add_hw_queue(struct amdgpu_mes *mes,
mes_add_queue_pkt.gds_size = input->queue_size;
return mes_v12_0_submit_pkt_and_poll_completion(mes,
AMDGPU_MES_SCHED_PIPE,
&mes_add_queue_pkt, sizeof(mes_add_queue_pkt),
offsetof(union MESAPI__ADD_QUEUE, api_status));
}
@@ -330,6 +345,7 @@ static int mes_v12_0_remove_hw_queue(struct amdgpu_mes *mes,
mes_remove_queue_pkt.gang_context_addr = input->gang_context_addr;
return mes_v12_0_submit_pkt_and_poll_completion(mes,
AMDGPU_MES_SCHED_PIPE,
&mes_remove_queue_pkt, sizeof(mes_remove_queue_pkt),
offsetof(union MESAPI__REMOVE_QUEUE, api_status));
}
@@ -338,6 +354,7 @@ static int mes_v12_0_map_legacy_queue(struct amdgpu_mes *mes,
struct mes_map_legacy_queue_input *input)
{
union MESAPI__ADD_QUEUE mes_add_queue_pkt;
int pipe;
memset(&mes_add_queue_pkt, 0, sizeof(mes_add_queue_pkt));
@@ -354,7 +371,12 @@ static int mes_v12_0_map_legacy_queue(struct amdgpu_mes *mes,
convert_to_mes_queue_type(input->queue_type);
mes_add_queue_pkt.map_legacy_kq = 1;
return mes_v12_0_submit_pkt_and_poll_completion(mes,
if (mes->adev->enable_uni_mes)
pipe = AMDGPU_MES_KIQ_PIPE;
else
pipe = AMDGPU_MES_SCHED_PIPE;
return mes_v12_0_submit_pkt_and_poll_completion(mes, pipe,
&mes_add_queue_pkt, sizeof(mes_add_queue_pkt),
offsetof(union MESAPI__ADD_QUEUE, api_status));
}
@@ -363,6 +385,7 @@ static int mes_v12_0_unmap_legacy_queue(struct amdgpu_mes *mes,
struct mes_unmap_legacy_queue_input *input)
{
union MESAPI__REMOVE_QUEUE mes_remove_queue_pkt;
int pipe;
memset(&mes_remove_queue_pkt, 0, sizeof(mes_remove_queue_pkt));
@@ -387,7 +410,12 @@ static int mes_v12_0_unmap_legacy_queue(struct amdgpu_mes *mes,
convert_to_mes_queue_type(input->queue_type);
}
return mes_v12_0_submit_pkt_and_poll_completion(mes,
if (mes->adev->enable_uni_mes)
pipe = AMDGPU_MES_KIQ_PIPE;
else
pipe = AMDGPU_MES_SCHED_PIPE;
return mes_v12_0_submit_pkt_and_poll_completion(mes, pipe,
&mes_remove_queue_pkt, sizeof(mes_remove_queue_pkt),
offsetof(union MESAPI__REMOVE_QUEUE, api_status));
}
@@ -404,7 +432,7 @@ static int mes_v12_0_resume_gang(struct amdgpu_mes *mes,
return 0;
}
static int mes_v12_0_query_sched_status(struct amdgpu_mes *mes)
static int mes_v12_0_query_sched_status(struct amdgpu_mes *mes, int pipe)
{
union MESAPI__QUERY_MES_STATUS mes_status_pkt;
@@ -414,7 +442,7 @@ static int mes_v12_0_query_sched_status(struct amdgpu_mes *mes)
mes_status_pkt.header.opcode = MES_SCH_API_QUERY_SCHEDULER_STATUS;
mes_status_pkt.header.dwsize = API_FRAME_SIZE_IN_DWORDS;
return mes_v12_0_submit_pkt_and_poll_completion(mes,
return mes_v12_0_submit_pkt_and_poll_completion(mes, pipe,
&mes_status_pkt, sizeof(mes_status_pkt),
offsetof(union MESAPI__QUERY_MES_STATUS, api_status));
}
@@ -423,6 +451,7 @@ static int mes_v12_0_misc_op(struct amdgpu_mes *mes,
struct mes_misc_op_input *input)
{
union MESAPI__MISC misc_pkt;
int pipe;
memset(&misc_pkt, 0, sizeof(misc_pkt));
@@ -475,12 +504,17 @@ static int mes_v12_0_misc_op(struct amdgpu_mes *mes,
return -EINVAL;
}
return mes_v12_0_submit_pkt_and_poll_completion(mes,
if (mes->adev->enable_uni_mes)
pipe = AMDGPU_MES_KIQ_PIPE;
else
pipe = AMDGPU_MES_SCHED_PIPE;
return mes_v12_0_submit_pkt_and_poll_completion(mes, pipe,
&misc_pkt, sizeof(misc_pkt),
offsetof(union MESAPI__MISC, api_status));
}
static int mes_v12_0_set_hw_resources_1(struct amdgpu_mes *mes)
static int mes_v12_0_set_hw_resources_1(struct amdgpu_mes *mes, int pipe)
{
union MESAPI_SET_HW_RESOURCES_1 mes_set_hw_res_1_pkt;
@@ -491,12 +525,12 @@ static int mes_v12_0_set_hw_resources_1(struct amdgpu_mes *mes)
mes_set_hw_res_1_pkt.header.dwsize = API_FRAME_SIZE_IN_DWORDS;
mes_set_hw_res_1_pkt.mes_kiq_unmap_timeout = 100;
return mes_v12_0_submit_pkt_and_poll_completion(mes,
return mes_v12_0_submit_pkt_and_poll_completion(mes, pipe,
&mes_set_hw_res_1_pkt, sizeof(mes_set_hw_res_1_pkt),
offsetof(union MESAPI_SET_HW_RESOURCES_1, api_status));
}
static int mes_v12_0_set_hw_resources(struct amdgpu_mes *mes)
static int mes_v12_0_set_hw_resources(struct amdgpu_mes *mes, int pipe)
{
int i;
struct amdgpu_device *adev = mes->adev;
@@ -508,27 +542,33 @@ static int mes_v12_0_set_hw_resources(struct amdgpu_mes *mes)
mes_set_hw_res_pkt.header.opcode = MES_SCH_API_SET_HW_RSRC;
mes_set_hw_res_pkt.header.dwsize = API_FRAME_SIZE_IN_DWORDS;
mes_set_hw_res_pkt.vmid_mask_mmhub = mes->vmid_mask_mmhub;
mes_set_hw_res_pkt.vmid_mask_gfxhub = mes->vmid_mask_gfxhub;
mes_set_hw_res_pkt.gds_size = adev->gds.gds_size;
mes_set_hw_res_pkt.paging_vmid = 0;
mes_set_hw_res_pkt.g_sch_ctx_gpu_mc_ptr = mes->sch_ctx_gpu_addr;
if (pipe == AMDGPU_MES_SCHED_PIPE) {
mes_set_hw_res_pkt.vmid_mask_mmhub = mes->vmid_mask_mmhub;
mes_set_hw_res_pkt.vmid_mask_gfxhub = mes->vmid_mask_gfxhub;
mes_set_hw_res_pkt.gds_size = adev->gds.gds_size;
mes_set_hw_res_pkt.paging_vmid = 0;
for (i = 0; i < MAX_COMPUTE_PIPES; i++)
mes_set_hw_res_pkt.compute_hqd_mask[i] =
mes->compute_hqd_mask[i];
for (i = 0; i < MAX_GFX_PIPES; i++)
mes_set_hw_res_pkt.gfx_hqd_mask[i] =
mes->gfx_hqd_mask[i];
for (i = 0; i < MAX_SDMA_PIPES; i++)
mes_set_hw_res_pkt.sdma_hqd_mask[i] =
mes->sdma_hqd_mask[i];
for (i = 0; i < AMD_PRIORITY_NUM_LEVELS; i++)
mes_set_hw_res_pkt.aggregated_doorbells[i] =
mes->aggregated_doorbells[i];
}
mes_set_hw_res_pkt.g_sch_ctx_gpu_mc_ptr =
mes->sch_ctx_gpu_addr[pipe];
mes_set_hw_res_pkt.query_status_fence_gpu_mc_ptr =
mes->query_status_fence_gpu_addr;
for (i = 0; i < MAX_COMPUTE_PIPES; i++)
mes_set_hw_res_pkt.compute_hqd_mask[i] =
mes->compute_hqd_mask[i];
for (i = 0; i < MAX_GFX_PIPES; i++)
mes_set_hw_res_pkt.gfx_hqd_mask[i] = mes->gfx_hqd_mask[i];
for (i = 0; i < MAX_SDMA_PIPES; i++)
mes_set_hw_res_pkt.sdma_hqd_mask[i] = mes->sdma_hqd_mask[i];
for (i = 0; i < AMD_PRIORITY_NUM_LEVELS; i++)
mes_set_hw_res_pkt.aggregated_doorbells[i] =
mes->aggregated_doorbells[i];
mes->query_status_fence_gpu_addr[pipe];
for (i = 0; i < 5; i++) {
mes_set_hw_res_pkt.gc_base[i] = adev->reg_offset[GC_HWIP][0][i];
@@ -556,7 +596,7 @@ static int mes_v12_0_set_hw_resources(struct amdgpu_mes *mes)
mes_set_hw_res_pkt.event_intr_history_gpu_mc_ptr = mes->event_log_gpu_addr;
}
return mes_v12_0_submit_pkt_and_poll_completion(mes,
return mes_v12_0_submit_pkt_and_poll_completion(mes, pipe,
&mes_set_hw_res_pkt, sizeof(mes_set_hw_res_pkt),
offsetof(union MESAPI_SET_HW_RESOURCES, api_status));
}
@@ -734,16 +774,11 @@ static void mes_v12_0_enable(struct amdgpu_device *adev, bool enable)
if (enable) {
data = RREG32_SOC15(GC, 0, regCP_MES_CNTL);
data = REG_SET_FIELD(data, CP_MES_CNTL, MES_PIPE0_RESET, 1);
data = REG_SET_FIELD(data, CP_MES_CNTL, MES_PIPE1_RESET,
(!adev->enable_uni_mes && adev->enable_mes_kiq) ? 1 : 0);
data = REG_SET_FIELD(data, CP_MES_CNTL, MES_PIPE1_RESET, 1);
WREG32_SOC15(GC, 0, regCP_MES_CNTL, data);
mutex_lock(&adev->srbm_mutex);
for (pipe = 0; pipe < AMDGPU_MAX_MES_PIPES; pipe++) {
if ((!adev->enable_mes_kiq || adev->enable_uni_mes) &&
pipe == AMDGPU_MES_KIQ_PIPE)
continue;
soc21_grbm_select(adev, 3, pipe, 0, 0);
ucode_addr = adev->mes.uc_start_addr[pipe] >> 2;
@@ -757,8 +792,7 @@ static void mes_v12_0_enable(struct amdgpu_device *adev, bool enable)
/* unhalt MES and activate pipe0 */
data = REG_SET_FIELD(0, CP_MES_CNTL, MES_PIPE0_ACTIVE, 1);
data = REG_SET_FIELD(data, CP_MES_CNTL, MES_PIPE1_ACTIVE,
(!adev->enable_uni_mes && adev->enable_mes_kiq) ? 1 : 0);
data = REG_SET_FIELD(data, CP_MES_CNTL, MES_PIPE1_ACTIVE, 1);
WREG32_SOC15(GC, 0, regCP_MES_CNTL, data);
if (amdgpu_emu_mode)
@@ -774,8 +808,7 @@ static void mes_v12_0_enable(struct amdgpu_device *adev, bool enable)
data = REG_SET_FIELD(data, CP_MES_CNTL,
MES_INVALIDATE_ICACHE, 1);
data = REG_SET_FIELD(data, CP_MES_CNTL, MES_PIPE0_RESET, 1);
data = REG_SET_FIELD(data, CP_MES_CNTL, MES_PIPE1_RESET,
(!adev->enable_uni_mes && adev->enable_mes_kiq) ? 1 : 0);
data = REG_SET_FIELD(data, CP_MES_CNTL, MES_PIPE1_RESET, 1);
data = REG_SET_FIELD(data, CP_MES_CNTL, MES_HALT, 1);
WREG32_SOC15(GC, 0, regCP_MES_CNTL, data);
}
@@ -790,10 +823,6 @@ static void mes_v12_0_set_ucode_start_addr(struct amdgpu_device *adev)
mutex_lock(&adev->srbm_mutex);
for (pipe = 0; pipe < AMDGPU_MAX_MES_PIPES; pipe++) {
if ((!adev->enable_mes_kiq || adev->enable_uni_mes) &&
pipe == AMDGPU_MES_KIQ_PIPE)
continue;
/* me=3, queue=0 */
soc21_grbm_select(adev, 3, pipe, 0, 0);
@@ -1085,7 +1114,7 @@ static int mes_v12_0_kiq_enable_queue(struct amdgpu_device *adev)
return r;
}
kiq->pmf->kiq_map_queues(kiq_ring, &adev->mes.ring);
kiq->pmf->kiq_map_queues(kiq_ring, &adev->mes.ring[0]);
r = amdgpu_ring_test_ring(kiq_ring);
if (r) {
@@ -1101,14 +1130,12 @@ static int mes_v12_0_queue_init(struct amdgpu_device *adev,
struct amdgpu_ring *ring;
int r;
if (pipe == AMDGPU_MES_KIQ_PIPE)
if (!adev->enable_uni_mes && pipe == AMDGPU_MES_KIQ_PIPE)
ring = &adev->gfx.kiq[0].ring;
else if (pipe == AMDGPU_MES_SCHED_PIPE)
ring = &adev->mes.ring;
else
BUG();
ring = &adev->mes.ring[pipe];
if ((pipe == AMDGPU_MES_SCHED_PIPE) &&
if ((adev->enable_uni_mes || pipe == AMDGPU_MES_SCHED_PIPE) &&
(amdgpu_in_reset(adev) || adev->in_suspend)) {
*(ring->wptr_cpu_addr) = 0;
*(ring->rptr_cpu_addr) = 0;
@@ -1120,13 +1147,12 @@ static int mes_v12_0_queue_init(struct amdgpu_device *adev,
return r;
if (pipe == AMDGPU_MES_SCHED_PIPE) {
if (adev->enable_uni_mes) {
mes_v12_0_queue_init_register(ring);
} else {
if (adev->enable_uni_mes)
r = amdgpu_mes_map_legacy_queue(adev, ring);
else
r = mes_v12_0_kiq_enable_queue(adev);
if (r)
return r;
}
if (r)
return r;
} else {
mes_v12_0_queue_init_register(ring);
}
@@ -1146,25 +1172,29 @@ static int mes_v12_0_queue_init(struct amdgpu_device *adev,
return 0;
}
static int mes_v12_0_ring_init(struct amdgpu_device *adev)
static int mes_v12_0_ring_init(struct amdgpu_device *adev, int pipe)
{
struct amdgpu_ring *ring;
ring = &adev->mes.ring;
ring = &adev->mes.ring[pipe];
ring->funcs = &mes_v12_0_ring_funcs;
ring->me = 3;
ring->pipe = 0;
ring->pipe = pipe;
ring->queue = 0;
ring->ring_obj = NULL;
ring->use_doorbell = true;
ring->doorbell_index = adev->doorbell_index.mes_ring0 << 1;
ring->eop_gpu_addr = adev->mes.eop_gpu_addr[AMDGPU_MES_SCHED_PIPE];
ring->eop_gpu_addr = adev->mes.eop_gpu_addr[pipe];
ring->no_scheduler = true;
sprintf(ring->name, "mes_%d.%d.%d", ring->me, ring->pipe, ring->queue);
if (pipe == AMDGPU_MES_SCHED_PIPE)
ring->doorbell_index = adev->doorbell_index.mes_ring0 << 1;
else
ring->doorbell_index = adev->doorbell_index.mes_ring1 << 1;
return amdgpu_ring_init(adev, ring, 1024, NULL, 0,
AMDGPU_RING_PRIO_DEFAULT, NULL);
}
@@ -1178,7 +1208,7 @@ static int mes_v12_0_kiq_ring_init(struct amdgpu_device *adev)
ring = &adev->gfx.kiq[0].ring;
ring->me = 3;
ring->pipe = adev->enable_uni_mes ? 0 : 1;
ring->pipe = 1;
ring->queue = 0;
ring->adev = NULL;
@@ -1200,12 +1230,10 @@ static int mes_v12_0_mqd_sw_init(struct amdgpu_device *adev,
int r, mqd_size = sizeof(struct v12_compute_mqd);
struct amdgpu_ring *ring;
if (pipe == AMDGPU_MES_KIQ_PIPE)
if (!adev->enable_uni_mes && pipe == AMDGPU_MES_KIQ_PIPE)
ring = &adev->gfx.kiq[0].ring;
else if (pipe == AMDGPU_MES_SCHED_PIPE)
ring = &adev->mes.ring;
else
BUG();
ring = &adev->mes.ring[pipe];
if (ring->mqd_obj)
return 0;
@@ -1246,9 +1274,6 @@ static int mes_v12_0_sw_init(void *handle)
return r;
for (pipe = 0; pipe < AMDGPU_MAX_MES_PIPES; pipe++) {
if (!adev->enable_mes_kiq && pipe == AMDGPU_MES_KIQ_PIPE)
continue;
r = mes_v12_0_allocate_eop_buf(adev, pipe);
if (r)
return r;
@@ -1256,18 +1281,15 @@ static int mes_v12_0_sw_init(void *handle)
r = mes_v12_0_mqd_sw_init(adev, pipe);
if (r)
return r;
}
if (adev->enable_mes_kiq) {
r = mes_v12_0_kiq_ring_init(adev);
if (!adev->enable_uni_mes && pipe == AMDGPU_MES_KIQ_PIPE)
r = mes_v12_0_kiq_ring_init(adev);
else
r = mes_v12_0_ring_init(adev, pipe);
if (r)
return r;
}
r = mes_v12_0_ring_init(adev);
if (r)
return r;
return 0;
}
@@ -1276,9 +1298,6 @@ static int mes_v12_0_sw_fini(void *handle)
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
int pipe;
amdgpu_device_wb_free(adev, adev->mes.sch_ctx_offs);
amdgpu_device_wb_free(adev, adev->mes.query_status_fence_offs);
for (pipe = 0; pipe < AMDGPU_MAX_MES_PIPES; pipe++) {
kfree(adev->mes.mqd_backup[pipe]);
@@ -1286,18 +1305,21 @@ static int mes_v12_0_sw_fini(void *handle)
&adev->mes.eop_gpu_addr[pipe],
NULL);
amdgpu_ucode_release(&adev->mes.fw[pipe]);
if (adev->enable_uni_mes || pipe == AMDGPU_MES_SCHED_PIPE) {
amdgpu_bo_free_kernel(&adev->mes.ring[pipe].mqd_obj,
&adev->mes.ring[pipe].mqd_gpu_addr,
&adev->mes.ring[pipe].mqd_ptr);
amdgpu_ring_fini(&adev->mes.ring[pipe]);
}
}
amdgpu_bo_free_kernel(&adev->gfx.kiq[0].ring.mqd_obj,
&adev->gfx.kiq[0].ring.mqd_gpu_addr,
&adev->gfx.kiq[0].ring.mqd_ptr);
amdgpu_bo_free_kernel(&adev->mes.ring.mqd_obj,
&adev->mes.ring.mqd_gpu_addr,
&adev->mes.ring.mqd_ptr);
amdgpu_ring_fini(&adev->gfx.kiq[0].ring);
amdgpu_ring_fini(&adev->mes.ring);
if (!adev->enable_uni_mes) {
amdgpu_bo_free_kernel(&adev->gfx.kiq[0].ring.mqd_obj,
&adev->gfx.kiq[0].ring.mqd_gpu_addr,
&adev->gfx.kiq[0].ring.mqd_ptr);
amdgpu_ring_fini(&adev->gfx.kiq[0].ring);
}
if (adev->firmware.load_type == AMDGPU_FW_LOAD_DIRECT) {
mes_v12_0_free_ucode_buffers(adev, AMDGPU_MES_KIQ_PIPE);
@@ -1341,7 +1363,7 @@ static void mes_v12_0_kiq_dequeue_sched(struct amdgpu_device *adev)
soc21_grbm_select(adev, 0, 0, 0, 0);
mutex_unlock(&adev->srbm_mutex);
adev->mes.ring.sched.ready = false;
adev->mes.ring[0].sched.ready = false;
}
static void mes_v12_0_kiq_setting(struct amdgpu_ring *ring)
@@ -1362,10 +1384,10 @@ static int mes_v12_0_kiq_hw_init(struct amdgpu_device *adev)
{
int r = 0;
mes_v12_0_kiq_setting(&adev->gfx.kiq[0].ring);
if (adev->enable_uni_mes)
return mes_v12_0_hw_init(adev);
mes_v12_0_kiq_setting(&adev->mes.ring[AMDGPU_MES_KIQ_PIPE]);
else
mes_v12_0_kiq_setting(&adev->gfx.kiq[0].ring);
if (adev->firmware.load_type == AMDGPU_FW_LOAD_DIRECT) {
@@ -1392,6 +1414,14 @@ static int mes_v12_0_kiq_hw_init(struct amdgpu_device *adev)
if (r)
goto failure;
if (adev->enable_uni_mes) {
r = mes_v12_0_set_hw_resources(&adev->mes, AMDGPU_MES_KIQ_PIPE);
if (r)
goto failure;
mes_v12_0_set_hw_resources_1(&adev->mes, AMDGPU_MES_KIQ_PIPE);
}
r = mes_v12_0_hw_init(adev);
if (r)
goto failure;
@@ -1405,9 +1435,15 @@ failure:
static int mes_v12_0_kiq_hw_fini(struct amdgpu_device *adev)
{
if (adev->mes.ring.sched.ready) {
mes_v12_0_kiq_dequeue_sched(adev);
adev->mes.ring.sched.ready = false;
if (adev->mes.ring[0].sched.ready) {
if (adev->enable_uni_mes)
amdgpu_mes_unmap_legacy_queue(adev,
&adev->mes.ring[AMDGPU_MES_SCHED_PIPE],
RESET_QUEUES, 0, 0);
else
mes_v12_0_kiq_dequeue_sched(adev);
adev->mes.ring[0].sched.ready = false;
}
mes_v12_0_enable(adev, false);
@@ -1420,10 +1456,10 @@ static int mes_v12_0_hw_init(void *handle)
int r;
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
if (adev->mes.ring.sched.ready)
if (adev->mes.ring[0].sched.ready)
goto out;
if (!adev->enable_mes_kiq || adev->enable_uni_mes) {
if (!adev->enable_mes_kiq) {
if (adev->firmware.load_type == AMDGPU_FW_LOAD_DIRECT) {
r = mes_v12_0_load_microcode(adev,
AMDGPU_MES_SCHED_PIPE, true);
@@ -1443,23 +1479,23 @@ static int mes_v12_0_hw_init(void *handle)
mes_v12_0_enable(adev, true);
}
/* Enable the MES to handle doorbell ring on unmapped queue */
mes_v12_0_enable_unmapped_doorbell_handling(&adev->mes, true);
r = mes_v12_0_queue_init(adev, AMDGPU_MES_SCHED_PIPE);
if (r)
goto failure;
r = mes_v12_0_set_hw_resources(&adev->mes);
r = mes_v12_0_set_hw_resources(&adev->mes, AMDGPU_MES_SCHED_PIPE);
if (r)
goto failure;
if (adev->enable_uni_mes)
mes_v12_0_set_hw_resources_1(&adev->mes);
mes_v12_0_set_hw_resources_1(&adev->mes, AMDGPU_MES_SCHED_PIPE);
mes_v12_0_init_aggregated_doorbell(&adev->mes);
/* Enable the MES to handle doorbell ring on unmapped queue */
mes_v12_0_enable_unmapped_doorbell_handling(&adev->mes, true);
r = mes_v12_0_query_sched_status(&adev->mes);
r = mes_v12_0_query_sched_status(&adev->mes, AMDGPU_MES_SCHED_PIPE);
if (r) {
DRM_ERROR("MES is busy\n");
goto failure;
@@ -1472,7 +1508,7 @@ out:
* with MES enabled.
*/
adev->gfx.kiq[0].ring.sched.ready = false;
adev->mes.ring.sched.ready = true;
adev->mes.ring[0].sched.ready = true;
return 0;
@@ -1515,17 +1551,7 @@ static int mes_v12_0_early_init(void *handle)
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
int pipe, r;
if (adev->enable_uni_mes) {
r = amdgpu_mes_init_microcode(adev, AMDGPU_MES_SCHED_PIPE);
if (!r)
return 0;
adev->enable_uni_mes = false;
}
for (pipe = 0; pipe < AMDGPU_MAX_MES_PIPES; pipe++) {
if (!adev->enable_mes_kiq && pipe == AMDGPU_MES_KIQ_PIPE)
continue;
r = amdgpu_mes_init_microcode(adev, pipe);
if (r)
return r;
drivers/gpu/drm/amd/amdgpu/soc15d.h
+6
View File
@@ -76,6 +76,12 @@
((cond & 0xF) << 24) | \
((type & 0xF) << 28))
#define CP_PACKETJ_NOP 0x60000000
#define CP_PACKETJ_GET_REG(x) ((x) & 0x3FFFF)
#define CP_PACKETJ_GET_RES(x) (((x) >> 18) & 0x3F)
#define CP_PACKETJ_GET_COND(x) (((x) >> 24) & 0xF)
#define CP_PACKETJ_GET_TYPE(x) (((x) >> 28) & 0xF)
/* Packet 3 types */
#define PACKET3_NOP 0x10
#define PACKET3_SET_BASE 0x11
drivers/gpu/drm/amd/amdgpu/soc24.c
+2
View File
@@ -406,6 +406,7 @@ static int soc24_common_early_init(void *handle)
AMD_CG_SUPPORT_ATHUB_MGCG |
AMD_CG_SUPPORT_ATHUB_LS |
AMD_CG_SUPPORT_MC_MGCG |
AMD_CG_SUPPORT_HDP_SD |
AMD_CG_SUPPORT_MC_LS;
adev->pg_flags = AMD_PG_SUPPORT_VCN |
AMD_PG_SUPPORT_JPEG |
@@ -424,6 +425,7 @@ static int soc24_common_early_init(void *handle)
AMD_CG_SUPPORT_ATHUB_MGCG |
AMD_CG_SUPPORT_ATHUB_LS |
AMD_CG_SUPPORT_MC_MGCG |
AMD_CG_SUPPORT_HDP_SD |
AMD_CG_SUPPORT_MC_LS;
adev->pg_flags = AMD_PG_SUPPORT_VCN |
drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+3
View File
@@ -2893,6 +2893,9 @@ static int dm_suspend(void *handle)
hpd_rx_irq_work_suspend(dm);
if (adev->dm.dc->caps.ips_support)
dc_allow_idle_optimizations(adev->dm.dc, true);
dc_set_power_state(dm->dc, DC_ACPI_CM_POWER_STATE_D3);
dc_dmub_srv_set_power_state(dm->dc->ctx->dmub_srv, DC_ACPI_CM_POWER_STATE_D3);
drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c
+24 -9
View File
@@ -804,12 +804,25 @@ struct dsc_mst_fairness_params {
};
#if defined(CONFIG_DRM_AMD_DC_FP)
static int kbps_to_peak_pbn(int kbps)
static uint16_t get_fec_overhead_multiplier(struct dc_link *dc_link)
{
u8 link_coding_cap;
uint16_t fec_overhead_multiplier_x1000 = PBN_FEC_OVERHEAD_MULTIPLIER_8B_10B;
link_coding_cap = dc_link_dp_mst_decide_link_encoding_format(dc_link);
if (link_coding_cap == DP_128b_132b_ENCODING)
fec_overhead_multiplier_x1000 = PBN_FEC_OVERHEAD_MULTIPLIER_128B_132B;
return fec_overhead_multiplier_x1000;
}
static int kbps_to_peak_pbn(int kbps, uint16_t fec_overhead_multiplier_x1000)
{
u64 peak_kbps = kbps;
peak_kbps *= 1006;
peak_kbps = div_u64(peak_kbps, 1000);
peak_kbps *= fec_overhead_multiplier_x1000;
peak_kbps = div_u64(peak_kbps, 1000 * 1000);
return (int) DIV64_U64_ROUND_UP(peak_kbps * 64, (54 * 8 * 1000));
}
@@ -910,11 +923,12 @@ static int increase_dsc_bpp(struct drm_atomic_state *state,
int link_timeslots_used;
int fair_pbn_alloc;
int ret = 0;
uint16_t fec_overhead_multiplier_x1000 = get_fec_overhead_multiplier(dc_link);
for (i = 0; i < count; i++) {
if (vars[i + k].dsc_enabled) {
initial_slack[i] =
kbps_to_peak_pbn(params[i].bw_range.max_kbps) - vars[i + k].pbn;
kbps_to_peak_pbn(params[i].bw_range.max_kbps, fec_overhead_multiplier_x1000) - vars[i + k].pbn;
bpp_increased[i] = false;
remaining_to_increase += 1;
} else {
@@ -1010,6 +1024,7 @@ static int try_disable_dsc(struct drm_atomic_state *state,
int next_index;
int remaining_to_try = 0;
int ret;
uint16_t fec_overhead_multiplier_x1000 = get_fec_overhead_multiplier(dc_link);
for (i = 0; i < count; i++) {
if (vars[i + k].dsc_enabled
@@ -1039,7 +1054,7 @@ static int try_disable_dsc(struct drm_atomic_state *state,
if (next_index == -1)
break;
vars[next_index].pbn = kbps_to_peak_pbn(params[next_index].bw_range.stream_kbps);
vars[next_index].pbn = kbps_to_peak_pbn(params[next_index].bw_range.stream_kbps, fec_overhead_multiplier_x1000);
ret = drm_dp_atomic_find_time_slots(state,
params[next_index].port->mgr,
params[next_index].port,
@@ -1052,8 +1067,7 @@ static int try_disable_dsc(struct drm_atomic_state *state,
vars[next_index].dsc_enabled = false;
vars[next_index].bpp_x16 = 0;
} else {
vars[next_index].pbn = kbps_to_peak_pbn(
params[next_index].bw_range.max_kbps);
vars[next_index].pbn = kbps_to_peak_pbn(params[next_index].bw_range.stream_kbps, fec_overhead_multiplier_x1000);
ret = drm_dp_atomic_find_time_slots(state,
params[next_index].port->mgr,
params[next_index].port,
@@ -1082,6 +1096,7 @@ static int compute_mst_dsc_configs_for_link(struct drm_atomic_state *state,
int count = 0;
int i, k, ret;
bool debugfs_overwrite = false;
uint16_t fec_overhead_multiplier_x1000 = get_fec_overhead_multiplier(dc_link);
memset(params, 0, sizeof(params));
@@ -1146,7 +1161,7 @@ static int compute_mst_dsc_configs_for_link(struct drm_atomic_state *state,
/* Try no compression */
for (i = 0; i < count; i++) {
vars[i + k].aconnector = params[i].aconnector;
vars[i + k].pbn = kbps_to_peak_pbn(params[i].bw_range.stream_kbps);
vars[i + k].pbn = kbps_to_peak_pbn(params[i].bw_range.stream_kbps, fec_overhead_multiplier_x1000);
vars[i + k].dsc_enabled = false;
vars[i + k].bpp_x16 = 0;
ret = drm_dp_atomic_find_time_slots(state, params[i].port->mgr, params[i].port,
@@ -1165,7 +1180,7 @@ static int compute_mst_dsc_configs_for_link(struct drm_atomic_state *state,
/* Try max compression */
for (i = 0; i < count; i++) {
if (params[i].compression_possible && params[i].clock_force_enable != DSC_CLK_FORCE_DISABLE) {
vars[i + k].pbn = kbps_to_peak_pbn(params[i].bw_range.min_kbps);
vars[i + k].pbn = kbps_to_peak_pbn(params[i].bw_range.min_kbps, fec_overhead_multiplier_x1000);
vars[i + k].dsc_enabled = true;
vars[i + k].bpp_x16 = params[i].bw_range.min_target_bpp_x16;
ret = drm_dp_atomic_find_time_slots(state, params[i].port->mgr,
@@ -1173,7 +1188,7 @@ static int compute_mst_dsc_configs_for_link(struct drm_atomic_state *state,
if (ret < 0)
return ret;
} else {
vars[i + k].pbn = kbps_to_peak_pbn(params[i].bw_range.stream_kbps);
vars[i + k].pbn = kbps_to_peak_pbn(params[i].bw_range.stream_kbps, fec_overhead_multiplier_x1000);
vars[i + k].dsc_enabled = false;
vars[i + k].bpp_x16 = 0;
ret = drm_dp_atomic_find_time_slots(state, params[i].port->mgr,
drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.h
+3
View File
@@ -46,6 +46,9 @@
#define SYNAPTICS_CASCADED_HUB_ID 0x5A
#define IS_SYNAPTICS_CASCADED_PANAMERA(devName, data) ((IS_SYNAPTICS_PANAMERA(devName) && ((int)data[2] == SYNAPTICS_CASCADED_HUB_ID)) ? 1 : 0)
#define PBN_FEC_OVERHEAD_MULTIPLIER_8B_10B 1031
#define PBN_FEC_OVERHEAD_MULTIPLIER_128B_132B 1000
enum mst_msg_ready_type {
NONE_MSG_RDY_EVENT = 0,
DOWN_REP_MSG_RDY_EVENT = 1,
drivers/gpu/drm/amd/display/dc/hwss/dcn10/dcn10_hwseq.c
+2 -2
View File
@@ -3589,7 +3589,7 @@ void dcn10_set_cursor_position(struct pipe_ctx *pipe_ctx)
(int)hubp->curs_attr.width || pos_cpy.x
<= (int)hubp->curs_attr.width +
pipe_ctx->plane_state->src_rect.x) {
pos_cpy.x = temp_x + viewport_width;
pos_cpy.x = 2 * viewport_width - temp_x;
}
}
} else {
@@ -3682,7 +3682,7 @@ void dcn10_set_cursor_position(struct pipe_ctx *pipe_ctx)
(int)hubp->curs_attr.width || pos_cpy.x
<= (int)hubp->curs_attr.width +
pipe_ctx->plane_state->src_rect.x) {
pos_cpy.x = 2 * viewport_width - temp_x;
pos_cpy.x = temp_x + viewport_width;
}
}
} else {
drivers/gpu/drm/amd/display/dc/resource/dcn321/dcn321_resource.c
+3
View File
@@ -1778,6 +1778,9 @@ static bool dcn321_resource_construct(
dc->caps.color.mpc.ogam_rom_caps.hlg = 0;
dc->caps.color.mpc.ocsc = 1;
/* Use pipe context based otg sync logic */
dc->config.use_pipe_ctx_sync_logic = true;
dc->config.dc_mode_clk_limit_support = true;
dc->config.enable_windowed_mpo_odm = true;
/* read VBIOS LTTPR caps */
drivers/gpu/drm/amd/include/mes_v12_api_def.h
+6 -1
View File
@@ -97,6 +97,7 @@ enum MES_QUEUE_TYPE {
MES_QUEUE_TYPE_SDMA,
MES_QUEUE_TYPE_MAX,
MES_QUEUE_TYPE_SCHQ = MES_QUEUE_TYPE_MAX,
};
struct MES_API_STATUS {
@@ -242,8 +243,12 @@ union MESAPI_SET_HW_RESOURCES {
uint32_t send_write_data : 1;
uint32_t os_tdr_timeout_override : 1;
uint32_t use_rs64mem_for_proc_gang_ctx : 1;
uint32_t halt_on_misaligned_access : 1;
uint32_t use_add_queue_unmap_flag_addr : 1;
uint32_t enable_mes_sch_stb_log : 1;
uint32_t limit_single_process : 1;
uint32_t unmapped_doorbell_handling: 2;
uint32_t reserved : 15;
uint32_t reserved : 11;
};
uint32_t uint32_all;
};
drivers/gpu/drm/drm_panel_orientation_quirks.c
+12
View File
@@ -208,6 +208,18 @@ static const struct dmi_system_id orientation_data[] = {
DMI_MATCH(DMI_BOARD_NAME, "KUN"),
},
.driver_data = (void *)&lcd1600x2560_rightside_up,
}, { /* AYN Loki Max */
.matches = {
DMI_EXACT_MATCH(DMI_SYS_VENDOR, "ayn"),
DMI_EXACT_MATCH(DMI_PRODUCT_NAME, "Loki Max"),
},
.driver_data = (void *)&lcd1080x1920_leftside_up,
}, { /* AYN Loki Zero */
.matches = {
DMI_EXACT_MATCH(DMI_SYS_VENDOR, "ayn"),
DMI_EXACT_MATCH(DMI_PRODUCT_NAME, "Loki Zero"),
},
.driver_data = (void *)&lcd1080x1920_leftside_up,
}, { /* Chuwi HiBook (CWI514) */
.matches = {
DMI_MATCH(DMI_BOARD_VENDOR, "Hampoo"),
drivers/gpu/drm/mediatek/mtk_drm_drv.c
+2 -2
View File
@@ -539,8 +539,8 @@ static int mtk_drm_kms_init(struct drm_device *drm)
}
/* IGT will check if the cursor size is configured */
drm->mode_config.cursor_width = drm->mode_config.max_width;
drm->mode_config.cursor_height = drm->mode_config.max_height;
drm->mode_config.cursor_width = 512;
drm->mode_config.cursor_height = 512;
/* Use OVL device for all DMA memory allocations */
crtc = drm_crtc_from_index(drm, 0);
drivers/gpu/drm/rockchip/inno_hdmi.c
+1 -3
View File
@@ -279,7 +279,6 @@ static int inno_hdmi_upload_frame(struct drm_connector *connector,
const u8 *buffer, size_t len)
{
struct inno_hdmi *hdmi = connector_to_inno_hdmi(connector);
u8 packed_frame[HDMI_MAXIMUM_INFO_FRAME_SIZE];
ssize_t i;
if (type != HDMI_INFOFRAME_TYPE_AVI) {
@@ -291,8 +290,7 @@ static int inno_hdmi_upload_frame(struct drm_connector *connector,
inno_hdmi_disable_frame(connector, type);
for (i = 0; i < len; i++)
hdmi_writeb(hdmi, HDMI_CONTROL_PACKET_ADDR + i,
packed_frame[i]);
hdmi_writeb(hdmi, HDMI_CONTROL_PACKET_ADDR + i, buffer[i]);
return 0;
}

Some files were not shown because too many files have changed in this diff Show More