iv/linux
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Linux 4.4-rc2

Browse Source 
-----BEGIN PGP SIGNATURE-----
 Version: GnuPG v1
 
 iQEcBAABAgAGBQJWUmHZAAoJEHm+PkMAQRiGHtcH/RVRsn8re0WdRWYaTr9+Hknm
 CGlRJN4LKecttgYQ/2bS1QsDbt8usDPBiiYVopqGXQxPBmjyDAqPjsa+8VzCaVc6
 WA+9LDB+PcW28lD6BO+qSZCOAm7hHSZq7dtw9x658IqO+mI2mVeCybsAyunw2iWi
 Kf5q90wq6tIBXuT8YH9MXGrSCQw00NclbYeYwB9CmCt9hT/koEFBdl7uFUFitB+Q
 GSPTz5fXhgc5Lms85n7flZlrVKoQKmtDQe4/DvKZm+SjsATHU9ru89OxDBdS5gSG
 YcEIM4zc9tMjhs3GC9t6WXf6iFOdctum8HOhUoIN/+LVfeOMRRwAhRVqtGJ//Xw=
 =DCUg
 -----END PGP SIGNATURE-----

Merge tag 'v4.4-rc2' into patchwork

Linux 4.4-rc2

Several PCI media drivers got broken on Kernel 4.4-rc1, due to
pci_set_dma_mask() regressions. So, we need to add those fixes back
for the media drivers to work again.

* tag 'v4.4-rc2': (335 commits)
  Linux 4.4-rc2
  slab/slub: adjust kmem_cache_alloc_bulk API
  slub: add missing kmem cgroup support to kmem_cache_free_bulk
  slub: fix kmem cgroup bug in kmem_cache_alloc_bulk
  slub: optimize bulk slowpath free by detached freelist
  slub: support for bulk free with SLUB freelists
  parisc: Map kernel text and data on huge pages
  parisc: Add Huge Page and HUGETLBFS support
  parisc: Use long branch to do_syscall_trace_exit
  parisc: Increase initial kernel mapping to 32MB on 64bit kernel
  parisc: Initialize the fault vector earlier in the boot process.
  parisc: Add defines for Huge page support
  parisc: Drop unused MADV_xxxK_PAGES flags from asm/mman.h
  serial: export fsl8250_handle_irq
  serial: 8250_mid: Add missing dependency
  tty: audit: Fix audit source
  serial: etraxfs-uart: Fix crash
  serial: fsl_lpuart: Fix earlycon support
  bcm63xx_uart: Use the device name when registering an interrupt
  tty: Fix direct use of tty buffer work
  ...
This commit is contained in:
Mauro Carvalho Chehab 2015-11-23 14:16:58 -02:00
commit 10897dacea
334 changed files with 4198 additions and 5546 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
Documentation/IPMI.txt
View File

@ -587,7 +587,7 @@ used to control it:
modprobe ipmi_watchdog timeout=<t> pretimeout=<t> action=<action type>
preaction=<preaction type> preop=<preop type> start_now=x
nowayout=x ifnum_to_use=n
nowayout=x ifnum_to_use=n panic_wdt_timeout=<t>
ifnum_to_use specifies which interface the watchdog timer should use.
The default is -1, which means to pick the first one registered.
@ -597,7 +597,9 @@ is the amount of seconds before the reset that the pre-timeout panic will
occur (if pretimeout is zero, then pretimeout will not be enabled). Note
that the pretimeout is the time before the final timeout. So if the
timeout is 50 seconds and the pretimeout is 10 seconds, then the pretimeout
will occur in 40 second (10 seconds before the timeout).
will occur in 40 second (10 seconds before the timeout). The panic_wdt_timeout
is the value of timeout which is set on kernel panic, in order to let actions
such as kdump to occur during panic.
The action may be "reset", "power_cycle", or "power_off", and
specifies what to do when the timer times out, and defaults to
@ -634,6 +636,7 @@ for configuring the watchdog:
ipmi_watchdog.preop=<preop type>
ipmi_watchdog.start_now=x
ipmi_watchdog.nowayout=x
ipmi_watchdog.panic_wdt_timeout=<t>
The options are the same as the module parameter options.

1
Documentation/i2c/busses/i2c-i801
View File

@ -32,6 +32,7 @@ Supported adapters:
* Intel Sunrise Point-LP (PCH)
* Intel DNV (SOC)
* Intel Broxton (SOC)
* Intel Lewisburg (PCH)
Datasheets: Publicly available at the Intel website
On Intel Patsburg and later chipsets, both the normal host SMBus controller

3
Documentation/kernel-parameters.txt
View File

@ -1583,9 +1583,6 @@ bytes respectively. Such letter suffixes can also be entirely omitted.
hwp_only
Only load intel_pstate on systems which support
hardware P state control (HWP) if available.
no_acpi
Don't use ACPI processor performance control objects
_PSS and _PPC specified limits.
intremap= [X86-64, Intel-IOMMU]
on enable Interrupt Remapping (default)

40
MAINTAINERS
View File

@ -2449,7 +2449,9 @@ F: drivers/firmware/broadcom/*
BROADCOM STB NAND FLASH DRIVER
M: Brian Norris <computersforpeace@gmail.com>
M: Kamal Dasu <kdasu.kdev@gmail.com>
L: linux-mtd@lists.infradead.org
L: bcm-kernel-feedback-list@broadcom.com
S: Maintained
F: drivers/mtd/nand/brcmnand/
@ -2546,7 +2548,7 @@ F: arch/c6x/
CACHEFILES: FS-CACHE BACKEND FOR CACHING ON MOUNTED FILESYSTEMS
M: David Howells <dhowells@redhat.com>
L: linux-cachefs@redhat.com
L: linux-cachefs@redhat.com (moderated for non-subscribers)
S: Supported
F: Documentation/filesystems/caching/cachefiles.txt
F: fs/cachefiles/
@ -2929,10 +2931,9 @@ S: Maintained
F: drivers/platform/x86/compal-laptop.c
CONEXANT ACCESSRUNNER USB DRIVER
M: Simon Arlott <cxacru@fire.lp0.eu>
L: accessrunner-general@lists.sourceforge.net
W: http://accessrunner.sourceforge.net/
S: Maintained
S: Orphan
F: drivers/usb/atm/cxacru.c
CONFIGFS
@ -4409,6 +4410,7 @@ K: fmc_d.*register
FPGA MANAGER FRAMEWORK
M: Alan Tull <atull@opensource.altera.com>
R: Moritz Fischer <moritz.fischer@ettus.com>
S: Maintained
F: drivers/fpga/
F: include/linux/fpga/fpga-mgr.h
@ -4559,7 +4561,7 @@ F: include/linux/frontswap.h
FS-CACHE: LOCAL CACHING FOR NETWORK FILESYSTEMS
M: David Howells <dhowells@redhat.com>
L: linux-cachefs@redhat.com
L: linux-cachefs@redhat.com (moderated for non-subscribers)
S: Supported
F: Documentation/filesystems/caching/
F: fs/fscache/
@ -5711,13 +5713,6 @@ M: Juanjo Ciarlante <jjciarla@raiz.uncu.edu.ar>
S: Maintained
F: net/ipv4/netfilter/ipt_MASQUERADE.c
IP1000A 10/100/1000 GIGABIT ETHERNET DRIVER
M: Francois Romieu <romieu@fr.zoreil.com>
M: Sorbica Shieh <sorbica@icplus.com.tw>
L: netdev@vger.kernel.org
S: Maintained
F: drivers/net/ethernet/icplus/ipg.*
IPATH DRIVER
M: Mike Marciniszyn <infinipath@intel.com>
L: linux-rdma@vger.kernel.org
@ -6924,13 +6919,21 @@ F: drivers/scsi/megaraid.*
F: drivers/scsi/megaraid/
MELLANOX ETHERNET DRIVER (mlx4_en)
M: Amir Vadai <amirv@mellanox.com>
M: Eugenia Emantayev <eugenia@mellanox.com>
L: netdev@vger.kernel.org
S: Supported
W: http://www.mellanox.com
Q: http://patchwork.ozlabs.org/project/netdev/list/
F: drivers/net/ethernet/mellanox/mlx4/en_*
MELLANOX ETHERNET DRIVER (mlx5e)
M: Saeed Mahameed <saeedm@mellanox.com>
L: netdev@vger.kernel.org
S: Supported
W: http://www.mellanox.com
Q: http://patchwork.ozlabs.org/project/netdev/list/
F: drivers/net/ethernet/mellanox/mlx5/core/en_*
MELLANOX ETHERNET SWITCH DRIVERS
M: Jiri Pirko <jiri@mellanox.com>
M: Ido Schimmel <idosch@mellanox.com>
@ -7902,6 +7905,18 @@ S: Maintained
F: net/openvswitch/
F: include/uapi/linux/openvswitch.h
OPERATING PERFORMANCE POINTS (OPP)
M: Viresh Kumar <vireshk@kernel.org>
M: Nishanth Menon <nm@ti.com>
M: Stephen Boyd <sboyd@codeaurora.org>
L: linux-pm@vger.kernel.org
S: Maintained
T: git git://git.kernel.org/pub/scm/linux/kernel/git/vireshk/pm.git
F: drivers/base/power/opp/
F: include/linux/pm_opp.h
F: Documentation/power/opp.txt
F: Documentation/devicetree/bindings/opp/
OPL4 DRIVER
M: Clemens Ladisch <clemens@ladisch.de>
L: alsa-devel@alsa-project.org (moderated for non-subscribers)
@ -9315,7 +9330,6 @@ F: drivers/i2c/busses/i2c-designware-*
F: include/linux/platform_data/i2c-designware.h
SYNOPSYS DESIGNWARE MMC/SD/SDIO DRIVER
M: Seungwon Jeon <tgih.jun@samsung.com>
M: Jaehoon Chung <jh80.chung@samsung.com>
L: linux-mmc@vger.kernel.org
S: Maintained

2
Makefile
View File

@ -1,7 +1,7 @@
VERSION = 4
PATCHLEVEL = 4
SUBLEVEL = 0
EXTRAVERSION = -rc1
EXTRAVERSION = -rc2
NAME = Blurry Fish Butt
# *DOCUMENTATION*

16
arch/arm/boot/dts/imx27.dtsi
View File

@ -486,7 +486,10 @@
compatible = "fsl,imx27-usb";
reg = <0x10024000 0x200>;
interrupts = <56>;
clocks = <&clks IMX27_CLK_USB_IPG_GATE>;
clocks = <&clks IMX27_CLK_USB_IPG_GATE>,
<&clks IMX27_CLK_USB_AHB_GATE>,
<&clks IMX27_CLK_USB_DIV>;
clock-names = "ipg", "ahb", "per";
fsl,usbmisc = <&usbmisc 0>;
status = "disabled";
};
@ -495,7 +498,10 @@
compatible = "fsl,imx27-usb";
reg = <0x10024200 0x200>;
interrupts = <54>;
clocks = <&clks IMX27_CLK_USB_IPG_GATE>;
clocks = <&clks IMX27_CLK_USB_IPG_GATE>,
<&clks IMX27_CLK_USB_AHB_GATE>,
<&clks IMX27_CLK_USB_DIV>;
clock-names = "ipg", "ahb", "per";
fsl,usbmisc = <&usbmisc 1>;
dr_mode = "host";
status = "disabled";
@ -505,7 +511,10 @@
compatible = "fsl,imx27-usb";
reg = <0x10024400 0x200>;
interrupts = <55>;
clocks = <&clks IMX27_CLK_USB_IPG_GATE>;
clocks = <&clks IMX27_CLK_USB_IPG_GATE>,
<&clks IMX27_CLK_USB_AHB_GATE>,
<&clks IMX27_CLK_USB_DIV>;
clock-names = "ipg", "ahb", "per";
fsl,usbmisc = <&usbmisc 2>;
dr_mode = "host";
status = "disabled";
@ -515,7 +524,6 @@
#index-cells = <1>;
compatible = "fsl,imx27-usbmisc";
reg = <0x10024600 0x200>;
clocks = <&clks IMX27_CLK_USB_AHB_GATE>;
};
sahara2: sahara@10025000 {

2
arch/arm/net/bpf_jit_32.c
View File

@ -1061,7 +1061,7 @@ void bpf_jit_compile(struct bpf_prog *fp)
}
build_epilogue(&ctx);
flush_icache_range((u32)ctx.target, (u32)(ctx.target + ctx.idx));
flush_icache_range((u32)header, (u32)(ctx.target + ctx.idx));
#if __LINUX_ARM_ARCH__ < 7
if (ctx.imm_count)

2
arch/arm64/crypto/aes-ce-cipher.c
View File

@ -237,7 +237,7 @@ EXPORT_SYMBOL(ce_aes_setkey);
static struct crypto_alg aes_alg = {
.cra_name = "aes",
.cra_driver_name = "aes-ce",
.cra_priority = 300,
.cra_priority = 250,
.cra_flags = CRYPTO_ALG_TYPE_CIPHER,
.cra_blocksize = AES_BLOCK_SIZE,
.cra_ctxsize = sizeof(struct crypto_aes_ctx),

16
arch/arm64/include/asm/barrier.h
View File

@ -64,27 +64,31 @@ do { \
#define smp_load_acquire(p) \
({ \
typeof(*p) ___p1; \
union { typeof(*p) __val; char __c[1]; } __u; \
compiletime_assert_atomic_type(*p); \
switch (sizeof(*p)) { \
case 1: \
asm volatile ("ldarb %w0, %1" \
: "=r" (___p1) : "Q" (*p) : "memory"); \
: "=r" (*(__u8 *)__u.__c) \
: "Q" (*p) : "memory"); \
break; \
case 2: \
asm volatile ("ldarh %w0, %1" \
: "=r" (___p1) : "Q" (*p) : "memory"); \
: "=r" (*(__u16 *)__u.__c) \
: "Q" (*p) : "memory"); \
break; \
case 4: \
asm volatile ("ldar %w0, %1" \
: "=r" (___p1) : "Q" (*p) : "memory"); \
: "=r" (*(__u32 *)__u.__c) \
: "Q" (*p) : "memory"); \
break; \
case 8: \
asm volatile ("ldar %0, %1" \
: "=r" (___p1) : "Q" (*p) : "memory"); \
: "=r" (*(__u64 *)__u.__c) \
: "Q" (*p) : "memory"); \
break; \
} \
___p1; \
__u.__val; \
})
#define read_barrier_depends() do { } while(0)

3
arch/arm64/include/asm/compat.h
View File

@ -23,7 +23,6 @@
*/
#include <linux/types.h>
#include <linux/sched.h>
#include <linux/ptrace.h>
#define COMPAT_USER_HZ 100
#ifdef __AARCH64EB__
@ -234,7 +233,7 @@ static inline compat_uptr_t ptr_to_compat(void __user *uptr)
return (u32)(unsigned long)uptr;
}
#define compat_user_stack_pointer() (user_stack_pointer(current_pt_regs()))
#define compat_user_stack_pointer() (user_stack_pointer(task_pt_regs(current)))
static inline void __user *arch_compat_alloc_user_space(long len)
{

13
arch/arm64/include/asm/dma-mapping.h
View File

@ -18,7 +18,6 @@
#ifdef __KERNEL__
#include <linux/acpi.h>
#include <linux/types.h>
#include <linux/vmalloc.h>
@ -26,22 +25,16 @@
#include <asm/xen/hypervisor.h>
#define DMA_ERROR_CODE (~(dma_addr_t)0)
extern struct dma_map_ops *dma_ops;
extern struct dma_map_ops dummy_dma_ops;
static inline struct dma_map_ops *__generic_dma_ops(struct device *dev)
{
if (unlikely(!dev))
return dma_ops;
else if (dev->archdata.dma_ops)
if (dev && dev->archdata.dma_ops)
return dev->archdata.dma_ops;
else if (acpi_disabled)
return dma_ops;
/*
* When ACPI is enabled, if arch_set_dma_ops is not called,
* we will disable device DMA capability by setting it
* to dummy_dma_ops.
* We expect no ISA devices, and all other DMA masters are expected to
* have someone call arch_setup_dma_ops at device creation time.
*/
return &dummy_dma_ops;
}

2
arch/arm64/include/asm/mmu_context.h
View File

@ -101,7 +101,7 @@ static inline void cpu_set_default_tcr_t0sz(void)
#define destroy_context(mm) do { } while(0)
void check_and_switch_context(struct mm_struct *mm, unsigned int cpu);
#define init_new_context(tsk,mm) ({ atomic64_set(&mm->context.id, 0); 0; })
#define init_new_context(tsk,mm) ({ atomic64_set(&(mm)->context.id, 0); 0; })
/*
* This is called when "tsk" is about to enter lazy TLB mode.

1
arch/arm64/include/asm/pgtable.h
View File

@ -81,6 +81,7 @@ extern void __pgd_error(const char *file, int line, unsigned long val);
#define PAGE_KERNEL __pgprot(_PAGE_DEFAULT | PTE_PXN | PTE_UXN | PTE_DIRTY | PTE_WRITE)
#define PAGE_KERNEL_RO __pgprot(_PAGE_DEFAULT | PTE_PXN | PTE_UXN | PTE_DIRTY | PTE_RDONLY)
#define PAGE_KERNEL_ROX __pgprot(_PAGE_DEFAULT | PTE_UXN | PTE_DIRTY | PTE_RDONLY)
#define PAGE_KERNEL_EXEC __pgprot(_PAGE_DEFAULT | PTE_UXN | PTE_DIRTY | PTE_WRITE)
#define PAGE_KERNEL_EXEC_CONT __pgprot(_PAGE_DEFAULT | PTE_UXN | PTE_DIRTY | PTE_WRITE | PTE_CONT)

5
arch/arm64/kernel/cpuinfo.c
View File

@ -30,6 +30,7 @@
#include <linux/seq_file.h>
#include <linux/sched.h>
#include <linux/smp.h>
#include <linux/delay.h>
/*
* In case the boot CPU is hotpluggable, we record its initial state and
@ -112,6 +113,10 @@ static int c_show(struct seq_file *m, void *v)
*/
seq_printf(m, "processor\t: %d\n", i);
seq_printf(m, "BogoMIPS\t: %lu.%02lu\n",
loops_per_jiffy / (500000UL/HZ),
loops_per_jiffy / (5000UL/HZ) % 100);
/*
* Dump out the common processor features in a single line.
* Userspace should read the hwcaps with getauxval(AT_HWCAP)

14
arch/arm64/kernel/efi.c
View File

@ -224,6 +224,8 @@ static bool __init efi_virtmap_init(void)
{
efi_memory_desc_t *md;
init_new_context(NULL, &efi_mm);
for_each_efi_memory_desc(&memmap, md) {
u64 paddr, npages, size;
pgprot_t prot;
@ -254,7 +256,8 @@ static bool __init efi_virtmap_init(void)
else
prot = PAGE_KERNEL;
create_pgd_mapping(&efi_mm, paddr, md->virt_addr, size, prot);
create_pgd_mapping(&efi_mm, paddr, md->virt_addr, size,
__pgprot(pgprot_val(prot) | PTE_NG));
}
return true;
}
@ -329,14 +332,7 @@ core_initcall(arm64_dmi_init);
static void efi_set_pgd(struct mm_struct *mm)
{
if (mm == &init_mm)
cpu_set_reserved_ttbr0();
else
cpu_switch_mm(mm->pgd, mm);
local_flush_tlb_all();
if (icache_is_aivivt())
__local_flush_icache_all();
switch_mm(NULL, mm, NULL);
}
void efi_virtmap_load(void)

10
arch/arm64/kernel/suspend.c
View File

@ -1,3 +1,4 @@
#include <linux/ftrace.h>
#include <linux/percpu.h>
#include <linux/slab.h>
#include <asm/cacheflush.h>
@ -70,6 +71,13 @@ int cpu_suspend(unsigned long arg, int (*fn)(unsigned long))
*/
local_dbg_save(flags);
/*
* Function graph tracer state gets incosistent when the kernel
* calls functions that never return (aka suspend finishers) hence
* disable graph tracing during their execution.
*/
pause_graph_tracing();
/*
* mm context saved on the stack, it will be restored when
* the cpu comes out of reset through the identity mapped
@ -111,6 +119,8 @@ int cpu_suspend(unsigned long arg, int (*fn)(unsigned long))
hw_breakpoint_restore(NULL);
}
unpause_graph_tracing();
/*
* Restore pstate flags. OS lock and mdscr have been already
* restored, so from this point onwards, debugging is fully

35
arch/arm64/mm/dma-mapping.c
View File

@ -18,6 +18,7 @@
*/
#include <linux/gfp.h>
#include <linux/acpi.h>
#include <linux/export.h>
#include <linux/slab.h>
#include <linux/genalloc.h>
@ -28,9 +29,6 @@
#include <asm/cacheflush.h>
struct dma_map_ops *dma_ops;
EXPORT_SYMBOL(dma_ops);
static pgprot_t __get_dma_pgprot(struct dma_attrs *attrs, pgprot_t prot,
bool coherent)
{
@ -515,13 +513,7 @@ EXPORT_SYMBOL(dummy_dma_ops);
static int __init arm64_dma_init(void)
{
int ret;
dma_ops = &swiotlb_dma_ops;
ret = atomic_pool_init();
return ret;
return atomic_pool_init();
}
arch_initcall(arm64_dma_init);
@ -552,10 +544,14 @@ static void *__iommu_alloc_attrs(struct device *dev, size_t size,
{
bool coherent = is_device_dma_coherent(dev);
int ioprot = dma_direction_to_prot(DMA_BIDIRECTIONAL, coherent);
size_t iosize = size;
void *addr;
if (WARN(!dev, "cannot create IOMMU mapping for unknown device\n"))
return NULL;
size = PAGE_ALIGN(size);
/*
* Some drivers rely on this, and we probably don't want the
* possibility of stale kernel data being read by devices anyway.
@ -566,7 +562,7 @@ static void *__iommu_alloc_attrs(struct device *dev, size_t size,
struct page **pages;
pgprot_t prot = __get_dma_pgprot(attrs, PAGE_KERNEL, coherent);
pages = iommu_dma_alloc(dev, size, gfp, ioprot, handle,
pages = iommu_dma_alloc(dev, iosize, gfp, ioprot, handle,
flush_page);
if (!pages)
return NULL;
@ -574,7 +570,7 @@ static void *__iommu_alloc_attrs(struct device *dev, size_t size,
addr = dma_common_pages_remap(pages, size, VM_USERMAP, prot,
__builtin_return_address(0));
if (!addr)
iommu_dma_free(dev, pages, size, handle);
iommu_dma_free(dev, pages, iosize, handle);
} else {
struct page *page;
/*
@ -591,7 +587,7 @@ static void *__iommu_alloc_attrs(struct device *dev, size_t size,
if (!addr)
return NULL;
*handle = iommu_dma_map_page(dev, page, 0, size, ioprot);
*handle = iommu_dma_map_page(dev, page, 0, iosize, ioprot);
if (iommu_dma_mapping_error(dev, *handle)) {
if (coherent)
__free_pages(page, get_order(size));
@ -606,6 +602,9 @@ static void *__iommu_alloc_attrs(struct device *dev, size_t size,
static void __iommu_free_attrs(struct device *dev, size_t size, void *cpu_addr,
dma_addr_t handle, struct dma_attrs *attrs)
{
size_t iosize = size;
size = PAGE_ALIGN(size);
/*
* @cpu_addr will be one of 3 things depending on how it was allocated:
* - A remapped array of pages from iommu_dma_alloc(), for all
@ -617,17 +616,17 @@ static void __iommu_free_attrs(struct device *dev, size_t size, void *cpu_addr,
* Hence how dodgy the below logic looks...
*/
if (__in_atomic_pool(cpu_addr, size)) {
iommu_dma_unmap_page(dev, handle, size, 0, NULL);
iommu_dma_unmap_page(dev, handle, iosize, 0, NULL);
__free_from_pool(cpu_addr, size);
} else if (is_vmalloc_addr(cpu_addr)){
struct vm_struct *area = find_vm_area(cpu_addr);
if (WARN_ON(!area || !area->pages))
return;
iommu_dma_free(dev, area->pages, size, &handle);
iommu_dma_free(dev, area->pages, iosize, &handle);
dma_common_free_remap(cpu_addr, size, VM_USERMAP);
} else {
iommu_dma_unmap_page(dev, handle, size, 0, NULL);
iommu_dma_unmap_page(dev, handle, iosize, 0, NULL);
__free_pages(virt_to_page(cpu_addr), get_order(size));
}
}
@ -984,8 +983,8 @@ static void __iommu_setup_dma_ops(struct device *dev, u64 dma_base, u64 size,
void arch_setup_dma_ops(struct device *dev, u64 dma_base, u64 size,
struct iommu_ops *iommu, bool coherent)
{
if (!acpi_disabled && !dev->archdata.dma_ops)
dev->archdata.dma_ops = dma_ops;
if (!dev->archdata.dma_ops)
dev->archdata.dma_ops = &swiotlb_dma_ops;
dev->archdata.dma_coherent = coherent;
__iommu_setup_dma_ops(dev, dma_base, size, iommu);

14
arch/arm64/mm/mmu.c
View File

@ -362,8 +362,8 @@ static void __init __map_memblock(phys_addr_t start, phys_addr_t end)
* for now. This will get more fine grained later once all memory
* is mapped
*/
unsigned long kernel_x_start = round_down(__pa(_stext), SECTION_SIZE);
unsigned long kernel_x_end = round_up(__pa(__init_end), SECTION_SIZE);
unsigned long kernel_x_start = round_down(__pa(_stext), SWAPPER_BLOCK_SIZE);
unsigned long kernel_x_end = round_up(__pa(__init_end), SWAPPER_BLOCK_SIZE);
if (end < kernel_x_start) {
create_mapping(start, __phys_to_virt(start),
@ -451,18 +451,18 @@ static void __init fixup_executable(void)
{
#ifdef CONFIG_DEBUG_RODATA
/* now that we are actually fully mapped, make the start/end more fine grained */
if (!IS_ALIGNED((unsigned long)_stext, SECTION_SIZE)) {
if (!IS_ALIGNED((unsigned long)_stext, SWAPPER_BLOCK_SIZE)) {
unsigned long aligned_start = round_down(__pa(_stext),
SECTION_SIZE);
SWAPPER_BLOCK_SIZE);
create_mapping(aligned_start, __phys_to_virt(aligned_start),
__pa(_stext) - aligned_start,
PAGE_KERNEL);
}
if (!IS_ALIGNED((unsigned long)__init_end, SECTION_SIZE)) {
if (!IS_ALIGNED((unsigned long)__init_end, SWAPPER_BLOCK_SIZE)) {
unsigned long aligned_end = round_up(__pa(__init_end),
SECTION_SIZE);
SWAPPER_BLOCK_SIZE);
create_mapping(__pa(__init_end), (unsigned long)__init_end,
aligned_end - __pa(__init_end),
PAGE_KERNEL);
@ -475,7 +475,7 @@ void mark_rodata_ro(void)
{
create_mapping_late(__pa(_stext), (unsigned long)_stext,
(unsigned long)_etext - (unsigned long)_stext,
PAGE_KERNEL_EXEC | PTE_RDONLY);
PAGE_KERNEL_ROX);
}
#endif

48
arch/arm64/net/bpf_jit_comp.c
View File

@ -50,7 +50,7 @@ static const int bpf2a64[] = {
[BPF_REG_8] = A64_R(21),
[BPF_REG_9] = A64_R(22),
/* read-only frame pointer to access stack */
[BPF_REG_FP] = A64_FP,
[BPF_REG_FP] = A64_R(25),
/* temporary register for internal BPF JIT */
[TMP_REG_1] = A64_R(23),
[TMP_REG_2] = A64_R(24),
@ -155,18 +155,49 @@ static void build_prologue(struct jit_ctx *ctx)
stack_size += 4; /* extra for skb_copy_bits buffer */
stack_size = STACK_ALIGN(stack_size);
/*
* BPF prog stack layout
*
* high
* original A64_SP => 0:+-----+ BPF prologue
* |FP/LR|
* current A64_FP => -16:+-----+
* | ... | callee saved registers
* +-----+
* | | x25/x26
* BPF fp register => -80:+-----+
* | |
* | ... | BPF prog stack
* | |
* | |
* current A64_SP => +-----+
* | |
* | ... | Function call stack
* | |
* +-----+
* low
*
*/
/* Save FP and LR registers to stay align with ARM64 AAPCS */
emit(A64_PUSH(A64_FP, A64_LR, A64_SP), ctx);
emit(A64_MOV(1, A64_FP, A64_SP), ctx);
/* Save callee-saved register */
emit(A64_PUSH(r6, r7, A64_SP), ctx);
emit(A64_PUSH(r8, r9, A64_SP), ctx);
if (ctx->tmp_used)
emit(A64_PUSH(tmp1, tmp2, A64_SP), ctx);
/* Set up BPF stack */
emit(A64_SUB_I(1, A64_SP, A64_SP, stack_size), ctx);
/* Save fp (x25) and x26. SP requires 16 bytes alignment */
emit(A64_PUSH(fp, A64_R(26), A64_SP), ctx);
/* Set up frame pointer */
/* Set up BPF prog stack base register (x25) */
emit(A64_MOV(1, fp, A64_SP), ctx);
/* Set up function call stack */
emit(A64_SUB_I(1, A64_SP, A64_SP, stack_size), ctx);
/* Clear registers A and X */
emit_a64_mov_i64(ra, 0, ctx);
emit_a64_mov_i64(rx, 0, ctx);
@ -190,14 +221,17 @@ static void build_epilogue(struct jit_ctx *ctx)
/* We're done with BPF stack */
emit(A64_ADD_I(1, A64_SP, A64_SP, stack_size), ctx);
/* Restore fs (x25) and x26 */
emit(A64_POP(fp, A64_R(26), A64_SP), ctx);
/* Restore callee-saved register */
if (ctx->tmp_used)
emit(A64_POP(tmp1, tmp2, A64_SP), ctx);
emit(A64_POP(r8, r9, A64_SP), ctx);
emit(A64_POP(r6, r7, A64_SP), ctx);
/* Restore frame pointer */
emit(A64_MOV(1, fp, A64_SP), ctx);
/* Restore FP/LR registers */
emit(A64_POP(A64_FP, A64_LR, A64_SP), ctx);
/* Set return value */
emit(A64_MOV(1, A64_R(0), r0), ctx);
@ -758,7 +792,7 @@ void bpf_int_jit_compile(struct bpf_prog *prog)
if (bpf_jit_enable > 1)
bpf_jit_dump(prog->len, image_size, 2, ctx.image);
bpf_flush_icache(ctx.image, ctx.image + ctx.idx);
bpf_flush_icache(header, ctx.image + ctx.idx);
set_memory_ro((unsigned long)header, header->pages);
prog->bpf_func = (void *)ctx.image;

7
arch/mips/ath79/setup.c
View File

@ -216,9 +216,9 @@ void __init plat_mem_setup(void)
AR71XX_RESET_SIZE);
ath79_pll_base = ioremap_nocache(AR71XX_PLL_BASE,
AR71XX_PLL_SIZE);
ath79_detect_sys_type();
ath79_ddr_ctrl_init();
ath79_detect_sys_type();
if (mips_machtype != ATH79_MACH_GENERIC_OF)
detect_memory_region(0, ATH79_MEM_SIZE_MIN, ATH79_MEM_SIZE_MAX);
@ -281,3 +281,8 @@ MIPS_MACHINE(ATH79_MACH_GENERIC,
"Generic",
"Generic AR71XX/AR724X/AR913X based board",
ath79_generic_init);
MIPS_MACHINE(ATH79_MACH_GENERIC_OF,
"DTB",
"Generic AR71XX/AR724X/AR913X based board (DT)",
NULL);

2
arch/mips/boot/dts/qca/ar9132.dtsi
View File

@ -107,7 +107,7 @@
miscintc: interrupt-controller@18060010 {
compatible = "qca,ar9132-misc-intc",
"qca,ar7100-misc-intc";
reg = <0x18060010 0x4>;
reg = <0x18060010 0x8>;
interrupt-parent = <&cpuintc>;
interrupts = <6>;

3
arch/mips/include/asm/page.h
View File

@ -200,8 +200,9 @@ static inline int pfn_valid(unsigned long pfn)
{
/* avoid <linux/mm.h> include hell */
extern unsigned long max_mapnr;
unsigned long pfn_offset = ARCH_PFN_OFFSET;
return pfn >= ARCH_PFN_OFFSET && pfn < max_mapnr;
return pfn >= pfn_offset && pfn < max_mapnr;
}
#elif defined(CONFIG_SPARSEMEM)

3
arch/parisc/Kconfig
View File

@ -108,6 +108,9 @@ config PGTABLE_LEVELS
default 3 if 64BIT && PARISC_PAGE_SIZE_4KB
default 2
config SYS_SUPPORTS_HUGETLBFS
def_bool y if PA20
source "init/Kconfig"
source "kernel/Kconfig.freezer"

85
arch/parisc/include/asm/hugetlb.h Normal file
View File

@ -0,0 +1,85 @@
#ifndef _ASM_PARISC64_HUGETLB_H
#define _ASM_PARISC64_HUGETLB_H
#include <asm/page.h>
#include <asm-generic/hugetlb.h>
void set_huge_pte_at(struct mm_struct *mm, unsigned long addr,
pte_t *ptep, pte_t pte);
pte_t huge_ptep_get_and_clear(struct mm_struct *mm, unsigned long addr,
pte_t *ptep);
static inline int is_hugepage_only_range(struct mm_struct *mm,
unsigned long addr,
unsigned long len) {
return 0;
}
/*
* If the arch doesn't supply something else, assume that hugepage
* size aligned regions are ok without further preparation.
*/
static inline int prepare_hugepage_range(struct file *file,
unsigned long addr, unsigned long len)
{
if (len & ~HPAGE_MASK)
return -EINVAL;
if (addr & ~HPAGE_MASK)
return -EINVAL;
return 0;
}
static inline void hugetlb_free_pgd_range(struct mmu_gather *tlb,
unsigned long addr, unsigned long end,
unsigned long floor,
unsigned long ceiling)
{
free_pgd_range(tlb, addr, end, floor, ceiling);
}
static inline void huge_ptep_clear_flush(struct vm_area_struct *vma,
unsigned long addr, pte_t *ptep)
{
}
static inline int huge_pte_none(pte_t pte)
{
return pte_none(pte);
}
static inline pte_t huge_pte_wrprotect(pte_t pte)
{
return pte_wrprotect(pte);
}
static inline void huge_ptep_set_wrprotect(struct mm_struct *mm,
unsigned long addr, pte_t *ptep)
{
pte_t old_pte = *ptep;
set_huge_pte_at(mm, addr, ptep, pte_wrprotect(old_pte));
}
static inline int huge_ptep_set_access_flags(struct vm_area_struct *vma,
unsigned long addr, pte_t *ptep,
pte_t pte, int dirty)
{
int changed = !pte_same(*ptep, pte);
if (changed) {
set_huge_pte_at(vma->vm_mm, addr, ptep, pte);
flush_tlb_page(vma, addr);
}
return changed;
}
static inline pte_t huge_ptep_get(pte_t *ptep)
{
return *ptep;
}
static inline void arch_clear_hugepage_flags(struct page *page)
{
}
#endif /* _ASM_PARISC64_HUGETLB_H */

13
arch/parisc/include/asm/page.h
View File

@ -145,11 +145,22 @@ extern int npmem_ranges;
#endif /* CONFIG_DISCONTIGMEM */
#ifdef CONFIG_HUGETLB_PAGE
#define HPAGE_SHIFT 22 /* 4MB (is this fixed?) */
#define HPAGE_SHIFT PMD_SHIFT /* fixed for transparent huge pages */
#define HPAGE_SIZE ((1UL) << HPAGE_SHIFT)
#define HPAGE_MASK (~(HPAGE_SIZE - 1))
#define HUGETLB_PAGE_ORDER (HPAGE_SHIFT - PAGE_SHIFT)
#if defined(CONFIG_64BIT) && defined(CONFIG_PARISC_PAGE_SIZE_4KB)
# define REAL_HPAGE_SHIFT 20 /* 20 = 1MB */
# define _HUGE_PAGE_SIZE_ENCODING_DEFAULT _PAGE_SIZE_ENCODING_1M
#elif !defined(CONFIG_64BIT) && defined(CONFIG_PARISC_PAGE_SIZE_4KB)
# define REAL_HPAGE_SHIFT 22 /* 22 = 4MB */
# define _HUGE_PAGE_SIZE_ENCODING_DEFAULT _PAGE_SIZE_ENCODING_4M
#else
# define REAL_HPAGE_SHIFT 24 /* 24 = 16MB */
# define _HUGE_PAGE_SIZE_ENCODING_DEFAULT _PAGE_SIZE_ENCODING_16M
#endif
#endif /* CONFIG_HUGETLB_PAGE */
#define virt_addr_valid(kaddr) pfn_valid(__pa(kaddr) >> PAGE_SHIFT)

2
arch/parisc/include/asm/pgalloc.h
View File

@ -35,7 +35,7 @@ static inline pgd_t *pgd_alloc(struct mm_struct *mm)
PxD_FLAG_VALID |
PxD_FLAG_ATTACHED)
+ (__u32)(__pa((unsigned long)pgd) >> PxD_VALUE_SHIFT));
/* The first pmd entry also is marked with _PAGE_GATEWAY as
/* The first pmd entry also is marked with PxD_FLAG_ATTACHED as
* a signal that this pmd may not be freed */
__pgd_val_set(*pgd, PxD_FLAG_ATTACHED);
#endif

26
arch/parisc/include/asm/pgtable.h
View File

@ -83,7 +83,11 @@ static inline void purge_tlb_entries(struct mm_struct *mm, unsigned long addr)
printk("%s:%d: bad pgd %08lx.\n", __FILE__, __LINE__, (unsigned long)pgd_val(e))
/* This is the size of the initially mapped kernel memory */
#define KERNEL_INITIAL_ORDER 24 /* 0 to 1<<24 = 16MB */
#ifdef CONFIG_64BIT
#define KERNEL_INITIAL_ORDER 25 /* 1<<25 = 32MB */
#else
#define KERNEL_INITIAL_ORDER 24 /* 1<<24 = 16MB */
#endif
#define KERNEL_INITIAL_SIZE (1 << KERNEL_INITIAL_ORDER)
#if CONFIG_PGTABLE_LEVELS == 3
@ -167,7 +171,7 @@ static inline void purge_tlb_entries(struct mm_struct *mm, unsigned long addr)
#define _PAGE_NO_CACHE_BIT 24 /* (0x080) Uncached Page (U bit) */
#define _PAGE_ACCESSED_BIT 23 /* (0x100) Software: Page Accessed */
#define _PAGE_PRESENT_BIT 22 /* (0x200) Software: translation valid */
/* bit 21 was formerly the FLUSH bit but is now unused */
#define _PAGE_HPAGE_BIT 21 /* (0x400) Software: Huge Page */
#define _PAGE_USER_BIT 20 /* (0x800) Software: User accessible page */
/* N.B. The bits are defined in terms of a 32 bit word above, so the */
@ -194,6 +198,7 @@ static inline void purge_tlb_entries(struct mm_struct *mm, unsigned long addr)
#define _PAGE_NO_CACHE (1 << xlate_pabit(_PAGE_NO_CACHE_BIT))
#define _PAGE_ACCESSED (1 << xlate_pabit(_PAGE_ACCESSED_BIT))
#define _PAGE_PRESENT (1 << xlate_pabit(_PAGE_PRESENT_BIT))
#define _PAGE_HUGE (1 << xlate_pabit(_PAGE_HPAGE_BIT))
#define _PAGE_USER (1 << xlate_pabit(_PAGE_USER_BIT))
#define _PAGE_TABLE (_PAGE_PRESENT | _PAGE_READ | _PAGE_WRITE | _PAGE_DIRTY | _PAGE_ACCESSED)
@ -217,7 +222,7 @@ static inline void purge_tlb_entries(struct mm_struct *mm, unsigned long addr)
#define PxD_FLAG_VALID (1 << xlate_pabit(_PxD_VALID_BIT))
#define PxD_FLAG_MASK (0xf)
#define PxD_FLAG_SHIFT (4)
#define PxD_VALUE_SHIFT (8) /* (PAGE_SHIFT-PxD_FLAG_SHIFT) */
#define PxD_VALUE_SHIFT (PFN_PTE_SHIFT-PxD_FLAG_SHIFT)
#ifndef __ASSEMBLY__
@ -362,6 +367,18 @@ static inline pte_t pte_mkyoung(pte_t pte) { pte_val(pte) |= _PAGE_ACCESSED; ret
static inline pte_t pte_mkwrite(pte_t pte) { pte_val(pte) |= _PAGE_WRITE; return pte; }
static inline pte_t pte_mkspecial(pte_t pte) { return pte; }
/*
* Huge pte definitions.
*/
#ifdef CONFIG_HUGETLB_PAGE
#define pte_huge(pte) (pte_val(pte) & _PAGE_HUGE)
#define pte_mkhuge(pte) (__pte(pte_val(pte) | _PAGE_HUGE))
#else
#define pte_huge(pte) (0)
#define pte_mkhuge(pte) (pte)
#endif
/*
* Conversion functions: convert a page and protection to a page entry,
* and a page entry and page directory to the page they refer to.
@ -410,8 +427,9 @@ static inline pte_t pte_modify(pte_t pte, pgprot_t newprot)
/* Find an entry in the second-level page table.. */
#if CONFIG_PGTABLE_LEVELS == 3
#define pmd_index(addr) (((addr) >> PMD_SHIFT) & (PTRS_PER_PMD - 1))
#define pmd_offset(dir,address) \
((pmd_t *) pgd_page_vaddr(*(dir)) + (((address)>>PMD_SHIFT) & (PTRS_PER_PMD-1)))
((pmd_t *) pgd_page_vaddr(*(dir)) + pmd_index(address))
#else
#define pmd_offset(dir,addr) ((pmd_t *) dir)
#endif

27
arch/parisc/include/asm/processor.h
View File

@ -192,33 +192,6 @@ void show_trace(struct task_struct *task, unsigned long *stack);
*/
typedef unsigned int elf_caddr_t;
#define start_thread_som(regs, new_pc, new_sp) do { \
unsigned long *sp = (unsigned long *)new_sp; \
__u32 spaceid = (__u32)current->mm->context; \
unsigned long pc = (unsigned long)new_pc; \
/* offset pc for priv. level */ \
pc |= 3; \
\
regs->iasq[0] = spaceid; \
regs->iasq[1] = spaceid; \
regs->iaoq[0] = pc; \
regs->iaoq[1] = pc + 4; \
regs->sr[2] = LINUX_GATEWAY_SPACE; \
regs->sr[3] = 0xffff; \
regs->sr[4] = spaceid; \
regs->sr[5] = spaceid; \
regs->sr[6] = spaceid; \
regs->sr[7] = spaceid; \
regs->gr[ 0] = USER_PSW; \
regs->gr[30] = ((new_sp)+63)&~63; \
regs->gr[31] = pc; \
\
get_user(regs->gr[26],&sp[0]); \
get_user(regs->gr[25],&sp[-1]); \
get_user(regs->gr[24],&sp[-2]); \
get_user(regs->gr[23],&sp[-3]); \
} while(0)
/* The ELF abi wants things done a "wee bit" differently than
* som does. Supporting this behavior here avoids
* having our own version of create_elf_tables.

10
arch/parisc/include/uapi/asm/mman.h
View File

@ -49,16 +49,6 @@
#define MADV_DONTFORK 10 /* don't inherit across fork */
#define MADV_DOFORK 11 /* do inherit across fork */
/* The range 12-64 is reserved for page size specification. */
#define MADV_4K_PAGES 12 /* Use 4K pages */
#define MADV_16K_PAGES 14 /* Use 16K pages */
#define MADV_64K_PAGES 16 /* Use 64K pages */
#define MADV_256K_PAGES 18 /* Use 256K pages */
#define MADV_1M_PAGES 20 /* Use 1 Megabyte pages */
#define MADV_4M_PAGES 22 /* Use 4 Megabyte pages */
#define MADV_16M_PAGES 24 /* Use 16 Megabyte pages */
#define MADV_64M_PAGES 26 /* Use 64 Megabyte pages */
#define MADV_MERGEABLE 65 /* KSM may merge identical pages */
#define MADV_UNMERGEABLE 66 /* KSM may not merge identical pages */

8
arch/parisc/kernel/asm-offsets.c
View File

@ -289,6 +289,14 @@ int main(void)
DEFINE(ASM_PTE_ENTRY_SIZE, PTE_ENTRY_SIZE);
DEFINE(ASM_PFN_PTE_SHIFT, PFN_PTE_SHIFT);
DEFINE(ASM_PT_INITIAL, PT_INITIAL);
BLANK();
/* HUGEPAGE_SIZE is only used in vmlinux.lds.S to align kernel text
* and kernel data on physical huge pages */
#ifdef CONFIG_HUGETLB_PAGE
DEFINE(HUGEPAGE_SIZE, 1UL << REAL_HPAGE_SHIFT);
#else
DEFINE(HUGEPAGE_SIZE, PAGE_SIZE);
#endif
BLANK();
DEFINE(EXCDATA_IP, offsetof(struct exception_data, fault_ip));
DEFINE(EXCDATA_SPACE, offsetof(struct exception_data, fault_space));

56
arch/parisc/kernel/entry.S
View File

@ -502,21 +502,38 @@
STREG \pte,0(\ptp)
.endm
/* We have (depending on the page size):
* - 38 to 52-bit Physical Page Number
* - 12 to 26-bit page offset
*/
/* bitshift difference between a PFN (based on kernel's PAGE_SIZE)
* to a CPU TLB 4k PFN (4k => 12 bits to shift) */
#define PAGE_ADD_SHIFT (PAGE_SHIFT-12)
#define PAGE_ADD_SHIFT (PAGE_SHIFT-12)
#define PAGE_ADD_HUGE_SHIFT (REAL_HPAGE_SHIFT-12)
/* Drop prot bits and convert to page addr for iitlbt and idtlbt */
.macro convert_for_tlb_insert20 pte
.macro convert_for_tlb_insert20 pte,tmp
#ifdef CONFIG_HUGETLB_PAGE
copy \pte,\tmp
extrd,u \tmp,(63-ASM_PFN_PTE_SHIFT)+(63-58)+PAGE_ADD_SHIFT,\
64-PAGE_SHIFT-PAGE_ADD_SHIFT,\pte
depdi _PAGE_SIZE_ENCODING_DEFAULT,63,\
(63-58)+PAGE_ADD_SHIFT,\pte
extrd,u,*= \tmp,_PAGE_HPAGE_BIT+32,1,%r0
depdi _HUGE_PAGE_SIZE_ENCODING_DEFAULT,63,\
(63-58)+PAGE_ADD_HUGE_SHIFT,\pte
#else /* Huge pages disabled */
extrd,u \pte,(63-ASM_PFN_PTE_SHIFT)+(63-58)+PAGE_ADD_SHIFT,\
64-PAGE_SHIFT-PAGE_ADD_SHIFT,\pte
depdi _PAGE_SIZE_ENCODING_DEFAULT,63,\
(63-58)+PAGE_ADD_SHIFT,\pte
#endif
.endm
/* Convert the pte and prot to tlb insertion values. How
* this happens is quite subtle, read below */
.macro make_insert_tlb spc,pte,prot
.macro make_insert_tlb spc,pte,prot,tmp
space_to_prot \spc \prot /* create prot id from space */
/* The following is the real subtlety. This is depositing
* T <-> _PAGE_REFTRAP
@ -553,7 +570,7 @@
depdi 1,12,1,\prot
/* Drop prot bits and convert to page addr for iitlbt and idtlbt */
convert_for_tlb_insert20 \pte
convert_for_tlb_insert20 \pte \tmp
.endm
/* Identical macro to make_insert_tlb above, except it
@ -646,17 +663,12 @@
/*
* Align fault_vector_20 on 4K boundary so that both
* fault_vector_11 and fault_vector_20 are on the
* same page. This is only necessary as long as we
* write protect the kernel text, which we may stop
* doing once we use large page translations to cover
* the static part of the kernel address space.
* Fault_vectors are architecturally required to be aligned on a 2K
* boundary
*/
.text
.align 4096
.align 2048
ENTRY(fault_vector_20)
/* First vector is invalid (0) */
@ -1147,7 +1159,7 @@ dtlb_miss_20w:
tlb_lock spc,ptp,pte,t0,t1,dtlb_check_alias_20w
update_accessed ptp,pte,t0,t1
make_insert_tlb spc,pte,prot
make_insert_tlb spc,pte,prot,t1
idtlbt pte,prot
@ -1173,7 +1185,7 @@ nadtlb_miss_20w:
tlb_lock spc,ptp,pte,t0,t1,nadtlb_check_alias_20w
update_accessed ptp,pte,t0,t1
make_insert_tlb spc,pte,prot
make_insert_tlb spc,pte,prot,t1
idtlbt pte,prot
@ -1267,7 +1279,7 @@ dtlb_miss_20:
tlb_lock spc,ptp,pte,t0,t1,dtlb_check_alias_20
update_accessed ptp,pte,t0,t1
make_insert_tlb spc,pte,prot
make_insert_tlb spc,pte,prot,t1
f_extend pte,t1
@ -1295,7 +1307,7 @@ nadtlb_miss_20:
tlb_lock spc,ptp,pte,t0,t1,nadtlb_check_alias_20
update_accessed ptp,pte,t0,t1
make_insert_tlb spc,pte,prot
make_insert_tlb spc,pte,prot,t1
f_extend pte,t1
@ -1404,7 +1416,7 @@ itlb_miss_20w:
tlb_lock spc,ptp,pte,t0,t1,itlb_fault
update_accessed ptp,pte,t0,t1
make_insert_tlb spc,pte,prot
make_insert_tlb spc,pte,prot,t1
iitlbt pte,prot
@ -1428,7 +1440,7 @@ naitlb_miss_20w:
tlb_lock spc,ptp,pte,t0,t1,naitlb_check_alias_20w
update_accessed ptp,pte,t0,t1
make_insert_tlb spc,pte,prot
make_insert_tlb spc,pte,prot,t1
iitlbt pte,prot
@ -1514,7 +1526,7 @@ itlb_miss_20:
tlb_lock spc,ptp,pte,t0,t1,itlb_fault
update_accessed ptp,pte,t0,t1
make_insert_tlb spc,pte,prot
make_insert_tlb spc,pte,prot,t1
f_extend pte,t1
@ -1534,7 +1546,7 @@ naitlb_miss_20:
tlb_lock spc,ptp,pte,t0,t1,naitlb_check_alias_20
update_accessed ptp,pte,t0,t1
make_insert_tlb spc,pte,prot
make_insert_tlb spc,pte,prot,t1
f_extend pte,t1
@ -1566,7 +1578,7 @@ dbit_trap_20w:
tlb_lock spc,ptp,pte,t0,t1,dbit_fault
update_dirty ptp,pte,t1
make_insert_tlb spc,pte,prot
make_insert_tlb spc,pte,prot,t1
idtlbt pte,prot
@ -1610,7 +1622,7 @@ dbit_trap_20:
tlb_lock spc,ptp,pte,t0,t1,dbit_fault
update_dirty ptp,pte,t1
make_insert_tlb spc,pte,prot
make_insert_tlb spc,pte,prot,t1
f_extend pte,t1

4
arch/parisc/kernel/head.S
View File

@ -69,7 +69,7 @@ $bss_loop:
stw,ma %arg2,4(%r1)
stw,ma %arg3,4(%r1)
/* Initialize startup VM. Just map first 8/16 MB of memory */
/* Initialize startup VM. Just map first 16/32 MB of memory */
load32 PA(swapper_pg_dir),%r4
mtctl %r4,%cr24 /* Initialize kernel root pointer */
mtctl %r4,%cr25 /* Initialize user root pointer */
@ -107,7 +107,7 @@ $bss_loop:
/* Now initialize the PTEs themselves. We use RWX for
* everything ... it will get remapped correctly later */
ldo 0+_PAGE_KERNEL_RWX(%r0),%r3 /* Hardwired 0 phys addr start */
ldi (1<<(KERNEL_INITIAL_ORDER-PAGE_SHIFT)),%r11 /* PFN count */
load32 (1<<(KERNEL_INITIAL_ORDER-PAGE_SHIFT)),%r11 /* PFN count */
load32 PA(pg0),%r1
$pgt_fill_loop:

14
arch/parisc/kernel/setup.c
View File

@ -130,7 +130,16 @@ void __init setup_arch(char **cmdline_p)
printk(KERN_INFO "The 32-bit Kernel has started...\n");
#endif
printk(KERN_INFO "Default page size is %dKB.\n", (int)(PAGE_SIZE / 1024));
printk(KERN_INFO "Kernel default page size is %d KB. Huge pages ",
(int)(PAGE_SIZE / 1024));
#ifdef CONFIG_HUGETLB_PAGE
printk(KERN_CONT "enabled with %d MB physical and %d MB virtual size",
1 << (REAL_HPAGE_SHIFT - 20), 1 << (HPAGE_SHIFT - 20));
#else
printk(KERN_CONT "disabled");
#endif
printk(KERN_CONT ".\n");
pdc_console_init();
@ -377,6 +386,7 @@ arch_initcall(parisc_init);
void start_parisc(void)
{
extern void start_kernel(void);
extern void early_trap_init(void);
int ret, cpunum;
struct pdc_coproc_cfg coproc_cfg;
@ -397,6 +407,8 @@ void start_parisc(void)
panic("must have an fpu to boot linux");
}
early_trap_init(); /* initialize checksum of fault_vector */
start_kernel();
// not reached
}

4
arch/parisc/kernel/syscall.S
View File

@ -369,7 +369,7 @@ tracesys_exit:
ldo -16(%r30),%r29 /* Reference param save area */
#endif
ldo TASK_REGS(%r1),%r26
bl do_syscall_trace_exit,%r2
BL do_syscall_trace_exit,%r2
STREG %r28,TASK_PT_GR28(%r1) /* save return value now */
ldo -THREAD_SZ_ALGN-FRAME_SIZE(%r30),%r1 /* get task ptr */
LDREG TI_TASK(%r1), %r1
@ -390,7 +390,7 @@ tracesys_sigexit:
#ifdef CONFIG_64BIT
ldo -16(%r30),%r29 /* Reference param save area */
#endif
bl do_syscall_trace_exit,%r2
BL do_syscall_trace_exit,%r2
ldo TASK_REGS(%r1),%r26
ldil L%syscall_exit_rfi,%r1

35
arch/parisc/kernel/traps.c
View File

@ -807,7 +807,7 @@ void notrace handle_interruption(int code, struct pt_regs *regs)
}
int __init check_ivt(void *iva)
void __init initialize_ivt(const void *iva)
{
extern u32 os_hpmc_size;
extern const u32 os_hpmc[];
@ -818,8 +818,8 @@ int __init check_ivt(void *iva)
u32 *hpmcp;
u32 length;
if (strcmp((char *)iva, "cows can fly"))
return -1;
if (strcmp((const char *)iva, "cows can fly"))
panic("IVT invalid");
ivap = (u32 *)iva;
@ -839,28 +839,23 @@ int __init check_ivt(void *iva)
check += ivap[i];
ivap[5] = -check;
return 0;
}
/* early_trap_init() is called before we set up kernel mappings and
* write-protect the kernel */
void __init early_trap_init(void)
{
extern const void fault_vector_20;
#ifndef CONFIG_64BIT
extern const void fault_vector_11;
extern const void fault_vector_11;
initialize_ivt(&fault_vector_11);
#endif
extern const void fault_vector_20;
initialize_ivt(&fault_vector_20);
}
void __init trap_init(void)
{
void *iva;
if (boot_cpu_data.cpu_type >= pcxu)
iva = (void *) &fault_vector_20;
else
#ifdef CONFIG_64BIT
panic("Can't boot 64-bit OS on PA1.1 processor!");
#else
iva = (void *) &fault_vector_11;
#endif
if (check_ivt(iva))
panic("IVT invalid");
}

9
arch/parisc/kernel/vmlinux.lds.S
View File

@ -60,7 +60,7 @@ SECTIONS
EXIT_DATA
}
PERCPU_SECTION(8)
. = ALIGN(PAGE_SIZE);
. = ALIGN(HUGEPAGE_SIZE);
__init_end = .;
/* freed after init ends here */
@ -116,7 +116,7 @@ SECTIONS
* that we can properly leave these
* as writable
*/
. = ALIGN(PAGE_SIZE);
. = ALIGN(HUGEPAGE_SIZE);
data_start = .;
EXCEPTION_TABLE(8)
@ -135,8 +135,11 @@ SECTIONS
_edata = .;
/* BSS */
BSS_SECTION(PAGE_SIZE, PAGE_SIZE, 8)
BSS_SECTION(PAGE_SIZE, PAGE_SIZE, PAGE_SIZE)
/* bootmap is allocated in setup_bootmem() directly behind bss. */
. = ALIGN(HUGEPAGE_SIZE);
_end = . ;
STABS_DEBUG

1
arch/parisc/mm/Makefile
View File

@ -3,3 +3,4 @@
#
obj-y := init.o fault.o ioremap.o
obj-$(CONFIG_HUGETLB_PAGE) += hugetlbpage.o

161
arch/parisc/mm/hugetlbpage.c Normal file
View File

@ -0,0 +1,161 @@
/*
* PARISC64 Huge TLB page support.
*
* This parisc implementation is heavily based on the SPARC and x86 code.
*
* Copyright (C) 2015 Helge Deller <deller@gmx.de>
*/
#include <linux/fs.h>
#include <linux/mm.h>
#include <linux/hugetlb.h>
#include <linux/pagemap.h>
#include <linux/sysctl.h>
#include <asm/mman.h>
#include <asm/pgalloc.h>
#include <asm/tlb.h>
#include <asm/tlbflush.h>
#include <asm/cacheflush.h>
#include <asm/mmu_context.h>
unsigned long
hugetlb_get_unmapped_area(struct file *file, unsigned long addr,
unsigned long len, unsigned long pgoff, unsigned long flags)
{
struct hstate *h = hstate_file(file);
if (len & ~huge_page_mask(h))
return -EINVAL;
if (len > TASK_SIZE)
return -ENOMEM;
if (flags & MAP_FIXED)
if (prepare_hugepage_range(file, addr, len))
return -EINVAL;
if (addr)
addr = ALIGN(addr, huge_page_size(h));
/* we need to make sure the colouring is OK */
return arch_get_unmapped_area(file, addr, len, pgoff, flags);
}
pte_t *huge_pte_alloc(struct mm_struct *mm,
unsigned long addr, unsigned long sz)
{
pgd_t *pgd;
pud_t *pud;
pmd_t *pmd;
pte_t *pte = NULL;
/* We must align the address, because our caller will run
* set_huge_pte_at() on whatever we return, which writes out
* all of the sub-ptes for the hugepage range. So we have
* to give it the first such sub-pte.
*/
addr &= HPAGE_MASK;
pgd = pgd_offset(mm, addr);
pud = pud_alloc(mm, pgd, addr);
if (pud) {
pmd = pmd_alloc(mm, pud, addr);
if (pmd)
pte = pte_alloc_map(mm, NULL, pmd, addr);
}
return pte;
}
pte_t *huge_pte_offset(struct mm_struct *mm, unsigned long addr)
{
pgd_t *pgd;
pud_t *pud;
pmd_t *pmd;
pte_t *pte = NULL;
addr &= HPAGE_MASK;
pgd = pgd_offset(mm, addr);
if (!pgd_none(*pgd)) {
pud = pud_offset(pgd, addr);
if (!pud_none(*pud)) {
pmd = pmd_offset(pud, addr);
if (!pmd_none(*pmd))
pte = pte_offset_map(pmd, addr);
}
}
return pte;
}
/* Purge data and instruction TLB entries. Must be called holding
* the pa_tlb_lock. The TLB purge instructions are slow on SMP
* machines since the purge must be broadcast to all CPUs.
*/
static inline void purge_tlb_entries_huge(struct mm_struct *mm, unsigned long addr)
{
int i;
/* We may use multiple physical huge pages (e.g. 2x1 MB) to emulate
* Linux standard huge pages (e.g. 2 MB) */
BUILD_BUG_ON(REAL_HPAGE_SHIFT > HPAGE_SHIFT);
addr &= HPAGE_MASK;
addr |= _HUGE_PAGE_SIZE_ENCODING_DEFAULT;
for (i = 0; i < (1 << (HPAGE_SHIFT-REAL_HPAGE_SHIFT)); i++) {
mtsp(mm->context, 1);
pdtlb(addr);
if (unlikely(split_tlb))
pitlb(addr);
addr += (1UL << REAL_HPAGE_SHIFT);
}
}
void set_huge_pte_at(struct mm_struct *mm, unsigned long addr,
pte_t *ptep, pte_t entry)
{
unsigned long addr_start;
int i;
addr &= HPAGE_MASK;
addr_start = addr;
for (i = 0; i < (1 << HUGETLB_PAGE_ORDER); i++) {
/* Directly write pte entry. We could call set_pte_at(mm, addr, ptep, entry)
* instead, but then we get double locking on pa_tlb_lock. */
*ptep = entry;
ptep++;
/* Drop the PAGE_SIZE/non-huge tlb entry */
purge_tlb_entries(mm, addr);
addr += PAGE_SIZE;
pte_val(entry) += PAGE_SIZE;
}
purge_tlb_entries_huge(mm, addr_start);
}
pte_t huge_ptep_get_and_clear(struct mm_struct *mm, unsigned long addr,
pte_t *ptep)
{
pte_t entry;
entry = *ptep;
set_huge_pte_at(mm, addr, ptep, __pte(0));
return entry;
}
int pmd_huge(pmd_t pmd)
{
return 0;
}
int pud_huge(pud_t pud)
{
return 0;
}

40
arch/parisc/mm/init.c
View File

@ -409,15 +409,11 @@ static void __init map_pages(unsigned long start_vaddr,
unsigned long vaddr;
unsigned long ro_start;
unsigned long ro_end;
unsigned long fv_addr;
unsigned long gw_addr;
extern const unsigned long fault_vector_20;
extern void * const linux_gateway_page;
unsigned long kernel_end;
ro_start = __pa((unsigned long)_text);
ro_end = __pa((unsigned long)&data_start);
fv_addr = __pa((unsigned long)&fault_vector_20) & PAGE_MASK;
gw_addr = __pa((unsigned long)&linux_gateway_page) & PAGE_MASK;
kernel_end = __pa((unsigned long)&_end);
end_paddr = start_paddr + size;
@ -475,24 +471,25 @@ static void __init map_pages(unsigned long start_vaddr,
for (tmp2 = start_pte; tmp2 < PTRS_PER_PTE; tmp2++, pg_table++) {
pte_t pte;
/*
* Map the fault vector writable so we can
* write the HPMC checksum.
*/
if (force)
pte = __mk_pte(address, pgprot);
else if (parisc_text_address(vaddr) &&
address != fv_addr)
else if (parisc_text_address(vaddr)) {
pte = __mk_pte(address, PAGE_KERNEL_EXEC);
if (address >= ro_start && address < kernel_end)
pte = pte_mkhuge(pte);
}
else
#if defined(CONFIG_PARISC_PAGE_SIZE_4KB)
if (address >= ro_start && address < ro_end
&& address != fv_addr
&& address != gw_addr)
pte = __mk_pte(address, PAGE_KERNEL_RO);
else
if (address >= ro_start && address < ro_end) {
pte = __mk_pte(address, PAGE_KERNEL_EXEC);
pte = pte_mkhuge(pte);
} else
#endif
{
pte = __mk_pte(address, pgprot);
if (address >= ro_start && address < kernel_end)
pte = pte_mkhuge(pte);
}
if (address >= end_paddr) {
if (force)
@ -536,15 +533,12 @@ void free_initmem(void)
/* force the kernel to see the new TLB entries */
__flush_tlb_range(0, init_begin, init_end);
/* Attempt to catch anyone trying to execute code here
* by filling the page with BRK insns.
*/
memset((void *)init_begin, 0x00, init_end - init_begin);
/* finally dump all the instructions which were cached, since the
* pages are no-longer executable */
flush_icache_range(init_begin, init_end);
free_initmem_default(-1);
free_initmem_default(POISON_FREE_INITMEM);
/* set up a new led state on systems shipped LED State panel */
pdc_chassis_send_status(PDC_CHASSIS_DIRECT_BCOMPLETE);
@ -728,8 +722,8 @@ static void __init pagetable_init(void)
unsigned long size;
start_paddr = pmem_ranges[range].start_pfn << PAGE_SHIFT;
end_paddr = start_paddr + (pmem_ranges[range].pages << PAGE_SHIFT);
size = pmem_ranges[range].pages << PAGE_SHIFT;
end_paddr = start_paddr + size;
map_pages((unsigned long)__va(start_paddr), start_paddr,
size, PAGE_KERNEL, 0);

1
arch/powerpc/include/asm/systbl.h
View File

@ -382,3 +382,4 @@ COMPAT_SYS(shmat)
SYSCALL(shmdt)
SYSCALL(shmget)
COMPAT_SYS(shmctl)
SYSCALL(mlock2)

2
arch/powerpc/include/asm/unistd.h
View File

@ -12,7 +12,7 @@
#include <uapi/asm/unistd.h>
#define __NR_syscalls 378
#define __NR_syscalls 379
#define __NR__exit __NR_exit
#define NR_syscalls __NR_syscalls

1
arch/powerpc/include/uapi/asm/unistd.h
View File

@ -400,5 +400,6 @@
#define __NR_shmdt 375
#define __NR_shmget 376
#define __NR_shmctl 377
#define __NR_mlock2 378
#endif /* _UAPI_ASM_POWERPC_UNISTD_H_ */

1
arch/s390/include/asm/cio.h
View File

@ -312,6 +312,7 @@ extern void css_schedule_reprobe(void);
extern void reipl_ccw_dev(struct ccw_dev_id *id);
struct cio_iplinfo {
u8 ssid;
u16 devno;
int is_qdio;
};

13
arch/s390/include/asm/elf.h
View File

@ -206,9 +206,16 @@ do { \
} while (0)
#endif /* CONFIG_COMPAT */
extern unsigned long mmap_rnd_mask;
#define STACK_RND_MASK (test_thread_flag(TIF_31BIT) ? 0x7ff : mmap_rnd_mask)
/*
* Cache aliasing on the latest machines calls for a mapping granularity
* of 512KB. For 64-bit processes use a 512KB alignment and a randomization
* of up to 1GB. For 31-bit processes the virtual address space is limited,
* use no alignment and limit the randomization to 8MB.
*/
#define BRK_RND_MASK (is_32bit_task() ? 0x7ffUL : 0x3ffffUL)
#define MMAP_RND_MASK (is_32bit_task() ? 0x7ffUL : 0x3ff80UL)
#define MMAP_ALIGN_MASK (is_32bit_task() ? 0 : 0x7fUL)
#define STACK_RND_MASK MMAP_RND_MASK
#define ARCH_DLINFO \
do { \

3
arch/s390/include/asm/ipl.h
View File

@ -64,7 +64,8 @@ struct ipl_block_fcp {
struct ipl_block_ccw {
u8 reserved1[84];
u8 reserved2[2];
u16 reserved2 : 13;
u8 ssid : 3;
u16 devno;
u8 vm_flags;
u8 reserved3[3];

4
arch/s390/include/asm/pci_dma.h
View File

@ -195,5 +195,7 @@ void zpci_dma_exit_device(struct zpci_dev *);
void dma_free_seg_table(unsigned long);
unsigned long *dma_alloc_cpu_table(void);
void dma_cleanup_tables(unsigned long *);
void dma_update_cpu_trans(unsigned long *, void *, dma_addr_t, int);
unsigned long *dma_walk_cpu_trans(unsigned long *rto, dma_addr_t dma_addr);
void dma_update_cpu_trans(unsigned long *entry, void *page_addr, int flags);
#endif

6
arch/s390/include/asm/trace/diag.h
View File

@ -19,7 +19,7 @@
#define TRACE_INCLUDE_PATH asm/trace
#define TRACE_INCLUDE_FILE diag
TRACE_EVENT(diagnose,
TRACE_EVENT(s390_diagnose,
TP_PROTO(unsigned short nr),
TP_ARGS(nr),
TP_STRUCT__entry(
@ -32,9 +32,9 @@ TRACE_EVENT(diagnose,
);
#ifdef CONFIG_TRACEPOINTS
void trace_diagnose_norecursion(int diag_nr);
void trace_s390_diagnose_norecursion(int diag_nr);
#else
static inline void trace_diagnose_norecursion(int diag_nr) { }
static inline void trace_s390_diagnose_norecursion(int diag_nr) { }
#endif
#endif /* _TRACE_S390_DIAG_H */

19
arch/s390/include/uapi/asm/unistd.h
View File

@ -192,14 +192,14 @@
#define __NR_set_tid_address 252
#define __NR_fadvise64 253
#define __NR_timer_create 254
#define __NR_timer_settime (__NR_timer_create+1)
#define __NR_timer_gettime (__NR_timer_create+2)
#define __NR_timer_getoverrun (__NR_timer_create+3)
#define __NR_timer_delete (__NR_timer_create+4)
#define __NR_clock_settime (__NR_timer_create+5)
#define __NR_clock_gettime (__NR_timer_create+6)
#define __NR_clock_getres (__NR_timer_create+7)
#define __NR_clock_nanosleep (__NR_timer_create+8)
#define __NR_timer_settime 255
#define __NR_timer_gettime 256
#define __NR_timer_getoverrun 257
#define __NR_timer_delete 258
#define __NR_clock_settime 259
#define __NR_clock_gettime 260
#define __NR_clock_getres 261
#define __NR_clock_nanosleep 262
/* Number 263 is reserved for vserver */
#define __NR_statfs64 265
#define __NR_fstatfs64 266
@ -309,7 +309,8 @@
#define __NR_recvfrom 371
#define __NR_recvmsg 372
#define __NR_shutdown 373
#define NR_syscalls 374
#define __NR_mlock2 374
#define NR_syscalls 375
/*
* There are some system calls that are not present on 64 bit, some

1
arch/s390/kernel/compat_wrapper.c
View File

@ -176,3 +176,4 @@ COMPAT_SYSCALL_WRAP4(accept4, int, fd, struct sockaddr __user *, upeer_sockaddr,
COMPAT_SYSCALL_WRAP3(getsockname, int, fd, struct sockaddr __user *, usockaddr, int __user *, usockaddr_len);
COMPAT_SYSCALL_WRAP3(getpeername, int, fd, struct sockaddr __user *, usockaddr, int __user *, usockaddr_len);
COMPAT_SYSCALL_WRAP6(sendto, int, fd, void __user *, buff, size_t, len, unsigned int, flags, struct sockaddr __user *, addr, int, addr_len);
COMPAT_SYSCALL_WRAP3(mlock2, unsigned long, start, size_t, len, int, flags);

4
arch/s390/kernel/diag.c
View File

@ -121,14 +121,14 @@ device_initcall(show_diag_stat_init);
void diag_stat_inc(enum diag_stat_enum nr)
{
this_cpu_inc(diag_stat.counter[nr]);
trace_diagnose(diag_map[nr].code);
trace_s390_diagnose(diag_map[nr].code);
}
EXPORT_SYMBOL(diag_stat_inc);
void diag_stat_inc_norecursion(enum diag_stat_enum nr)
{
this_cpu_inc(diag_stat.counter[nr]);
trace_diagnose_norecursion(diag_map[nr].code);
trace_s390_diagnose_norecursion(diag_map[nr].code);
}
EXPORT_SYMBOL(diag_stat_inc_norecursion);

95
arch/s390/kernel/head.S
View File

@ -26,6 +26,7 @@
#include <asm/asm-offsets.h>
#include <asm/thread_info.h>
#include <asm/page.h>
#include <asm/ptrace.h>
#define ARCH_OFFSET 4
@ -59,19 +60,6 @@ __HEAD
.long 0x020006e0,0x20000050
.org 0x200
#
# subroutine to set architecture mode
#
.Lsetmode:
mvi __LC_AR_MODE_ID,1 # set esame flag
slr %r0,%r0 # set cpuid to zero
lhi %r1,2 # mode 2 = esame (dump)
sigp %r1,%r0,0x12 # switch to esame mode
bras %r13,0f
.fill 16,4,0x0
0: lmh %r0,%r15,0(%r13) # clear high-order half of gprs
sam31 # switch to 31 bit addressing mode
br %r14
#
# subroutine to wait for end I/O
@ -159,7 +147,14 @@ __HEAD
.long 0x02200050,0x00000000
iplstart:
bas %r14,.Lsetmode # Immediately switch to 64 bit mode
mvi __LC_AR_MODE_ID,1 # set esame flag
slr %r0,%r0 # set cpuid to zero
lhi %r1,2 # mode 2 = esame (dump)
sigp %r1,%r0,0x12 # switch to esame mode
bras %r13,0f
.fill 16,4,0x0
0: lmh %r0,%r15,0(%r13) # clear high-order half of gprs
sam31 # switch to 31 bit addressing mode
lh %r1,0xb8 # test if subchannel number
bct %r1,.Lnoload # is valid
l %r1,0xb8 # load ipl subchannel number
@ -268,71 +263,6 @@ iplstart:
.align 8
.Lcpuid:.fill 8,1,0
#
# SALIPL loader support. Based on a patch by Rob van der Heij.
# This entry point is called directly from the SALIPL loader and
# doesn't need a builtin ipl record.
#
.org 0x800
ENTRY(start)
stm %r0,%r15,0x07b0 # store registers
bas %r14,.Lsetmode # Immediately switch to 64 bit mode
basr %r12,%r0
.base:
l %r11,.parm
l %r8,.cmd # pointer to command buffer
ltr %r9,%r9 # do we have SALIPL parameters?
bp .sk8x8
mvc 0(64,%r8),0x00b0 # copy saved registers
xc 64(240-64,%r8),0(%r8) # remainder of buffer
tr 0(64,%r8),.lowcase
b .gotr
.sk8x8:
mvc 0(240,%r8),0(%r9) # copy iplparms into buffer
.gotr:
slr %r0,%r0
st %r0,INITRD_SIZE+ARCH_OFFSET-PARMAREA(%r11)
st %r0,INITRD_START+ARCH_OFFSET-PARMAREA(%r11)
j startup # continue with startup
.cmd: .long COMMAND_LINE # address of command line buffer
.parm: .long PARMAREA
.lowcase:
.byte 0x00,0x01,0x02,0x03,0x04,0x05,0x06,0x07
.byte 0x08,0x09,0x0a,0x0b,0x0c,0x0d,0x0e,0x0f
.byte 0x10,0x11,0x12,0x13,0x14,0x15,0x16,0x17
.byte 0x18,0x19,0x1a,0x1b,0x1c,0x1d,0x1e,0x1f
.byte 0x20,0x21,0x22,0x23,0x24,0x25,0x26,0x27
.byte 0x28,0x29,0x2a,0x2b,0x2c,0x2d,0x2e,0x2f
.byte 0x30,0x31,0x32,0x33,0x34,0x35,0x36,0x37
.byte 0x38,0x39,0x3a,0x3b,0x3c,0x3d,0x3e,0x3f
.byte 0x40,0x41,0x42,0x43,0x44,0x45,0x46,0x47
.byte 0x48,0x49,0x4a,0x4b,0x4c,0x4d,0x4e,0x4f
.byte 0x50,0x51,0x52,0x53,0x54,0x55,0x56,0x57
.byte 0x58,0x59,0x5a,0x5b,0x5c,0x5d,0x5e,0x5f
.byte 0x60,0x61,0x62,0x63,0x64,0x65,0x66,0x67
.byte 0x68,0x69,0x6a,0x6b,0x6c,0x6d,0x6e,0x6f
.byte 0x70,0x71,0x72,0x73,0x74,0x75,0x76,0x77
.byte 0x78,0x79,0x7a,0x7b,0x7c,0x7d,0x7e,0x7f
.byte 0x80,0x81,0x82,0x83,0x84,0x85,0x86,0x87
.byte 0x88,0x89,0x8a,0x8b,0x8c,0x8d,0x8e,0x8f
.byte 0x90,0x91,0x92,0x93,0x94,0x95,0x96,0x97
.byte 0x98,0x99,0x9a,0x9b,0x9c,0x9d,0x9e,0x9f
.byte 0xa0,0xa1,0xa2,0xa3,0xa4,0xa5,0xa6,0xa7
.byte 0xa8,0xa9,0xaa,0xab,0xac,0xad,0xae,0xaf
.byte 0xb0,0xb1,0xb2,0xb3,0xb4,0xb5,0xb6,0xb7
.byte 0xb8,0xb9,0xba,0xbb,0xbc,0xbd,0xbe,0xbf
.byte 0xc0,0x81,0x82,0x83,0x84,0x85,0x86,0x87 # .abcdefg
.byte 0x88,0x89,0xca,0xcb,0xcc,0xcd,0xce,0xcf # hi
.byte 0xd0,0x91,0x92,0x93,0x94,0x95,0x96,0x97 # .jklmnop
.byte 0x98,0x99,0xda,0xdb,0xdc,0xdd,0xde,0xdf # qr
.byte 0xe0,0xe1,0xa2,0xa3,0xa4,0xa5,0xa6,0xa7 # ..stuvwx
.byte 0xa8,0xa9,0xea,0xeb,0xec,0xed,0xee,0xef # yz
.byte 0xf0,0xf1,0xf2,0xf3,0xf4,0xf5,0xf6,0xf7
.byte 0xf8,0xf9,0xfa,0xfb,0xfc,0xfd,0xfe,0xff
#
# startup-code at 0x10000, running in absolute addressing mode
# this is called either by the ipl loader or directly by PSW restart
@ -364,7 +294,7 @@ ENTRY(startup_kdump)
bras %r13,0f
.fill 16,4,0x0
0: lmh %r0,%r15,0(%r13) # clear high-order half of gprs
sam31 # switch to 31 bit addressing mode
sam64 # switch to 64 bit addressing mode
basr %r13,0 # get base
.LPG0:
xc 0x200(256),0x200 # partially clear lowcore
@ -395,7 +325,7 @@ ENTRY(startup_kdump)
jnz 1b
j 4f
2: l %r15,.Lstack-.LPG0(%r13)
ahi %r15,-96
ahi %r15,-STACK_FRAME_OVERHEAD
la %r2,.Lals_string-.LPG0(%r13)
l %r3,.Lsclp_print-.LPG0(%r13)
basr %r14,%r3
@ -429,8 +359,7 @@ ENTRY(startup_kdump)
.long 1, 0xc0000000
#endif
4:
/* Continue with 64bit startup code in head64.S */
sam64 # switch to 64 bit mode
/* Continue with startup code in head64.S */
jg startup_continue
.align 8

65
arch/s390/kernel/ipl.c
View File

@ -121,6 +121,7 @@ static char *dump_type_str(enum dump_type type)
* Must be in data section since the bss section
* is not cleared when these are accessed.
*/
static u8 ipl_ssid __attribute__((__section__(".data"))) = 0;
static u16 ipl_devno __attribute__((__section__(".data"))) = 0;
u32 ipl_flags __attribute__((__section__(".data"))) = 0;
@ -197,6 +198,33 @@ static ssize_t sys_##_prefix##_##_name##_show(struct kobject *kobj, \
return snprintf(page, PAGE_SIZE, _format, ##args); \
}
#define IPL_ATTR_CCW_STORE_FN(_prefix, _name, _ipl_blk) \
static ssize_t sys_##_prefix##_##_name##_store(struct kobject *kobj, \
struct kobj_attribute *attr, \
const char *buf, size_t len) \
{ \
unsigned long long ssid, devno; \
\
if (sscanf(buf, "0.%llx.%llx\n", &ssid, &devno) != 2) \
return -EINVAL; \
\
if (ssid > __MAX_SSID || devno > __MAX_SUBCHANNEL) \
return -EINVAL; \
\
_ipl_blk.ssid = ssid; \
_ipl_blk.devno = devno; \
return len; \
}
#define DEFINE_IPL_CCW_ATTR_RW(_prefix, _name, _ipl_blk) \
IPL_ATTR_SHOW_FN(_prefix, _name, "0.%x.%04x\n", \
_ipl_blk.ssid, _ipl_blk.devno); \
IPL_ATTR_CCW_STORE_FN(_prefix, _name, _ipl_blk); \
static struct kobj_attribute sys_##_prefix##_##_name##_attr = \
__ATTR(_name, (S_IRUGO | S_IWUSR), \
sys_##_prefix##_##_name##_show, \
sys_##_prefix##_##_name##_store) \
#define DEFINE_IPL_ATTR_RO(_prefix, _name, _format, _value) \
IPL_ATTR_SHOW_FN(_prefix, _name, _format, _value) \
static struct kobj_attribute sys_##_prefix##_##_name##_attr = \
@ -395,7 +423,7 @@ static ssize_t sys_ipl_device_show(struct kobject *kobj,
switch (ipl_info.type) {
case IPL_TYPE_CCW:
return sprintf(page, "0.0.%04x\n", ipl_devno);
return sprintf(page, "0.%x.%04x\n", ipl_ssid, ipl_devno);
case IPL_TYPE_FCP:
case IPL_TYPE_FCP_DUMP:
return sprintf(page, "0.0.%04x\n", ipl->ipl_info.fcp.devno);
@ -687,21 +715,14 @@ static ssize_t reipl_fcp_scpdata_write(struct file *filp, struct kobject *kobj,
struct bin_attribute *attr,
char *buf, loff_t off, size_t count)
{
size_t scpdata_len = count;
size_t padding;
size_t scpdata_len;
if (off < 0)
if (off)
return -EINVAL;
if (off >= DIAG308_SCPDATA_SIZE)
return -ENOSPC;
if (count > DIAG308_SCPDATA_SIZE - off)
count = DIAG308_SCPDATA_SIZE - off;
memcpy(reipl_block_fcp->ipl_info.fcp.scp_data, buf + off, count);
scpdata_len = off + count;
memcpy(reipl_block_fcp->ipl_info.fcp.scp_data, buf, count);
if (scpdata_len % 8) {
padding = 8 - (scpdata_len % 8);
memset(reipl_block_fcp->ipl_info.fcp.scp_data + scpdata_len,
@ -717,7 +738,7 @@ static ssize_t reipl_fcp_scpdata_write(struct file *filp, struct kobject *kobj,
}
static struct bin_attribute sys_reipl_fcp_scp_data_attr =
__BIN_ATTR(scp_data, (S_IRUGO | S_IWUSR), reipl_fcp_scpdata_read,
reipl_fcp_scpdata_write, PAGE_SIZE);
reipl_fcp_scpdata_write, DIAG308_SCPDATA_SIZE);
static struct bin_attribute *reipl_fcp_bin_attrs[] = {
&sys_reipl_fcp_scp_data_attr,
@ -814,9 +835,7 @@ static struct attribute_group reipl_fcp_attr_group = {
};
/* CCW reipl device attributes */
DEFINE_IPL_ATTR_RW(reipl_ccw, device, "0.0.%04llx\n", "0.0.%llx\n",
reipl_block_ccw->ipl_info.ccw.devno);
DEFINE_IPL_CCW_ATTR_RW(reipl_ccw, device, reipl_block_ccw->ipl_info.ccw);
/* NSS wrapper */
static ssize_t reipl_nss_loadparm_show(struct kobject *kobj,
@ -1056,8 +1075,8 @@ static void __reipl_run(void *unused)
switch (reipl_method) {
case REIPL_METHOD_CCW_CIO:
devid.ssid = reipl_block_ccw->ipl_info.ccw.ssid;
devid.devno = reipl_block_ccw->ipl_info.ccw.devno;
devid.ssid = 0;
reipl_ccw_dev(&devid);
break;
case REIPL_METHOD_CCW_VM:
@ -1192,6 +1211,7 @@ static int __init reipl_ccw_init(void)
reipl_block_ccw_init(reipl_block_ccw);
if (ipl_info.type == IPL_TYPE_CCW) {
reipl_block_ccw->ipl_info.ccw.ssid = ipl_ssid;
reipl_block_ccw->ipl_info.ccw.devno = ipl_devno;
reipl_block_ccw_fill_parms(reipl_block_ccw);
}
@ -1336,9 +1356,7 @@ static struct attribute_group dump_fcp_attr_group = {
};
/* CCW dump device attributes */
DEFINE_IPL_ATTR_RW(dump_ccw, device, "0.0.%04llx\n", "0.0.%llx\n",
dump_block_ccw->ipl_info.ccw.devno);
DEFINE_IPL_CCW_ATTR_RW(dump_ccw, device, dump_block_ccw->ipl_info.ccw);
static struct attribute *dump_ccw_attrs[] = {
&sys_dump_ccw_device_attr.attr,
@ -1418,8 +1436,8 @@ static void __dump_run(void *unused)
switch (dump_method) {
case DUMP_METHOD_CCW_CIO:
devid.ssid = dump_block_ccw->ipl_info.ccw.ssid;
devid.devno = dump_block_ccw->ipl_info.ccw.devno;
devid.ssid = 0;
reipl_ccw_dev(&devid);
break;
case DUMP_METHOD_CCW_VM:
@ -1939,14 +1957,14 @@ void __init setup_ipl(void)
ipl_info.type = get_ipl_type();
switch (ipl_info.type) {
case IPL_TYPE_CCW:
ipl_info.data.ccw.dev_id.ssid = ipl_ssid;
ipl_info.data.ccw.dev_id.devno = ipl_devno;
ipl_info.data.ccw.dev_id.ssid = 0;
break;
case IPL_TYPE_FCP:
case IPL_TYPE_FCP_DUMP:
ipl_info.data.fcp.dev_id.ssid = 0;
ipl_info.data.fcp.dev_id.devno =
IPL_PARMBLOCK_START->ipl_info.fcp.devno;
ipl_info.data.fcp.dev_id.ssid = 0;
ipl_info.data.fcp.wwpn = IPL_PARMBLOCK_START->ipl_info.fcp.wwpn;
ipl_info.data.fcp.lun = IPL_PARMBLOCK_START->ipl_info.fcp.lun;
break;
@ -1978,6 +1996,7 @@ void __init ipl_save_parameters(void)
if (cio_get_iplinfo(&iplinfo))
return;
ipl_ssid = iplinfo.ssid;
ipl_devno = iplinfo.devno;
ipl_flags |= IPL_DEVNO_VALID;
if (!iplinfo.is_qdio)

6
arch/s390/kernel/process.c
View File

@ -243,11 +243,7 @@ unsigned long arch_align_stack(unsigned long sp)
static inline unsigned long brk_rnd(void)
{
/* 8MB for 32bit, 1GB for 64bit */
if (is_32bit_task())
return (get_random_int() & 0x7ffUL) << PAGE_SHIFT;
else
return (get_random_int() & 0x3ffffUL) << PAGE_SHIFT;
return (get_random_int() & BRK_RND_MASK) << PAGE_SHIFT;
}
unsigned long arch_randomize_brk(struct mm_struct *mm)

2
arch/s390/kernel/sclp.c
View File

@ -21,7 +21,7 @@ static void _sclp_wait_int(void)
__ctl_load(cr0_new, 0, 0);
psw_ext_save = S390_lowcore.external_new_psw;
psw_mask = __extract_psw() & (PSW_MASK_EA | PSW_MASK_BA);
psw_mask = __extract_psw();
S390_lowcore.external_new_psw.mask = psw_mask;
psw_wait.mask = psw_mask | PSW_MASK_EXT | PSW_MASK_WAIT;
S390_lowcore.ext_int_code = 0;

3
arch/s390/kernel/setup.c
View File

@ -764,9 +764,6 @@ static int __init setup_hwcaps(void)
get_cpu_id(&cpu_id);
add_device_randomness(&cpu_id, sizeof(cpu_id));
switch (cpu_id.machine) {
case 0x9672:
strcpy(elf_platform, "g5");
break;
case 0x2064:
case 0x2066:
default: /* Use "z900" as default for 64 bit kernels. */

1
arch/s390/kernel/syscalls.S
View File

@ -382,3 +382,4 @@ SYSCALL(sys_sendmsg,compat_sys_sendmsg) /* 370 */
SYSCALL(sys_recvfrom,compat_sys_recvfrom)
SYSCALL(sys_recvmsg,compat_sys_recvmsg)
SYSCALL(sys_shutdown,sys_shutdown)
SYSCALL(sys_mlock2,compat_sys_mlock2)

6
arch/s390/kernel/trace.c
View File

@ -9,11 +9,11 @@
#define CREATE_TRACE_POINTS
#include <asm/trace/diag.h>
EXPORT_TRACEPOINT_SYMBOL(diagnose);
EXPORT_TRACEPOINT_SYMBOL(s390_diagnose);
static DEFINE_PER_CPU(unsigned int, diagnose_trace_depth);
void trace_diagnose_norecursion(int diag_nr)
void trace_s390_diagnose_norecursion(int diag_nr)
{
unsigned long flags;
unsigned int *depth;
@ -22,7 +22,7 @@ void trace_diagnose_norecursion(int diag_nr)
depth = this_cpu_ptr(&diagnose_trace_depth);
if (*depth == 0) {
(*depth)++;
trace_diagnose(diag_nr);
trace_s390_diagnose(diag_nr);
(*depth)--;
}
local_irq_restore(flags);

30
arch/s390/mm/init.c
View File

@ -48,37 +48,13 @@ EXPORT_SYMBOL(zero_page_mask);
static void __init setup_zero_pages(void)
{
struct cpuid cpu_id;
unsigned int order;
struct page *page;
int i;
get_cpu_id(&cpu_id);
switch (cpu_id.machine) {
case 0x9672: /* g5 */
case 0x2064: /* z900 */
case 0x2066: /* z900 */
case 0x2084: /* z990 */
case 0x2086: /* z990 */
case 0x2094: /* z9-109 */
case 0x2096: /* z9-109 */
order = 0;
break;
case 0x2097: /* z10 */
case 0x2098: /* z10 */
case 0x2817: /* z196 */
case 0x2818: /* z196 */
order = 2;
break;
case 0x2827: /* zEC12 */
case 0x2828: /* zEC12 */
order = 5;
break;
case 0x2964: /* z13 */
default:
order = 7;
break;
}
/* Latest machines require a mapping granularity of 512KB */
order = 7;
/* Limit number of empty zero pages for small memory sizes */
while (order > 2 && (totalram_pages >> 10) < (1UL << order))
order--;

60
arch/s390/mm/mmap.c
View File

@ -31,9 +31,6 @@
#include <linux/security.h>
#include <asm/pgalloc.h>
unsigned long mmap_rnd_mask;
static unsigned long mmap_align_mask;
static unsigned long stack_maxrandom_size(void)
{
if (!(current->flags & PF_RANDOMIZE))
@ -62,10 +59,7 @@ static inline int mmap_is_legacy(void)
unsigned long arch_mmap_rnd(void)
{
if (is_32bit_task())
return (get_random_int() & 0x7ff) << PAGE_SHIFT;
else
return (get_random_int() & mmap_rnd_mask) << PAGE_SHIFT;
return (get_random_int() & MMAP_RND_MASK) << PAGE_SHIFT;
}
static unsigned long mmap_base_legacy(unsigned long rnd)
@ -92,7 +86,6 @@ arch_get_unmapped_area(struct file *filp, unsigned long addr,
struct mm_struct *mm = current->mm;
struct vm_area_struct *vma;
struct vm_unmapped_area_info info;
int do_color_align;
if (len > TASK_SIZE - mmap_min_addr)
return -ENOMEM;
@ -108,15 +101,14 @@ arch_get_unmapped_area(struct file *filp, unsigned long addr,
return addr;
}
do_color_align = 0;
if (filp || (flags & MAP_SHARED))
do_color_align = !is_32bit_task();
info.flags = 0;
info.length = len;
info.low_limit = mm->mmap_base;
info.high_limit = TASK_SIZE;
info.align_mask = do_color_align ? (mmap_align_mask << PAGE_SHIFT) : 0;
if (filp || (flags & MAP_SHARED))
info.align_mask = MMAP_ALIGN_MASK << PAGE_SHIFT;
else
info.align_mask = 0;
info.align_offset = pgoff << PAGE_SHIFT;
return vm_unmapped_area(&info);
}
@ -130,7 +122,6 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0,
struct mm_struct *mm = current->mm;
unsigned long addr = addr0;
struct vm_unmapped_area_info info;
int do_color_align;
/* requested length too big for entire address space */
if (len > TASK_SIZE - mmap_min_addr)
@ -148,15 +139,14 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0,
return addr;
}
do_color_align = 0;
if (filp || (flags & MAP_SHARED))
do_color_align = !is_32bit_task();
info.flags = VM_UNMAPPED_AREA_TOPDOWN;
info.length = len;
info.low_limit = max(PAGE_SIZE, mmap_min_addr);
info.high_limit = mm->mmap_base;
info.align_mask = do_color_align ? (mmap_align_mask << PAGE_SHIFT) : 0;
if (filp || (flags & MAP_SHARED))
info.align_mask = MMAP_ALIGN_MASK << PAGE_SHIFT;
else
info.align_mask = 0;
info.align_offset = pgoff << PAGE_SHIFT;
addr = vm_unmapped_area(&info);
@ -254,35 +244,3 @@ void arch_pick_mmap_layout(struct mm_struct *mm)
mm->get_unmapped_area = s390_get_unmapped_area_topdown;
}
}
static int __init setup_mmap_rnd(void)
{
struct cpuid cpu_id;
get_cpu_id(&cpu_id);
switch (cpu_id.machine) {
case 0x9672:
case 0x2064:
case 0x2066:
case 0x2084:
case 0x2086:
case 0x2094:
case 0x2096:
case 0x2097:
case 0x2098:
case 0x2817:
case 0x2818:
case 0x2827:
case 0x2828:
mmap_rnd_mask = 0x7ffUL;
mmap_align_mask = 0UL;
break;
case 0x2964: /* z13 */
default:
mmap_rnd_mask = 0x3ff80UL;
mmap_align_mask = 0x7fUL;
break;
}
return 0;
}
early_initcall(setup_mmap_rnd);

84
arch/s390/pci/pci_dma.c
View File

@ -33,7 +33,7 @@ unsigned long *dma_alloc_cpu_table(void)
return NULL;
for (entry = table; entry < table + ZPCI_TABLE_ENTRIES; entry++)
*entry = ZPCI_TABLE_INVALID | ZPCI_TABLE_PROTECTED;
*entry = ZPCI_TABLE_INVALID;
return table;
}
@ -51,7 +51,7 @@ static unsigned long *dma_alloc_page_table(void)
return NULL;
for (entry = table; entry < table + ZPCI_PT_ENTRIES; entry++)
*entry = ZPCI_PTE_INVALID | ZPCI_TABLE_PROTECTED;
*entry = ZPCI_PTE_INVALID;
return table;
}
@ -95,7 +95,7 @@ static unsigned long *dma_get_page_table_origin(unsigned long *entry)
return pto;
}
static unsigned long *dma_walk_cpu_trans(unsigned long *rto, dma_addr_t dma_addr)
unsigned long *dma_walk_cpu_trans(unsigned long *rto, dma_addr_t dma_addr)
{
unsigned long *sto, *pto;
unsigned int rtx, sx, px;
@ -114,20 +114,10 @@ static unsigned long *dma_walk_cpu_trans(unsigned long *rto, dma_addr_t dma_addr
return &pto[px];
}
void dma_update_cpu_trans(unsigned long *dma_table, void *page_addr,
dma_addr_t dma_addr, int flags)
void dma_update_cpu_trans(unsigned long *entry, void *page_addr, int flags)
{
unsigned long *entry;
entry = dma_walk_cpu_trans(dma_table, dma_addr);
if (!entry) {
WARN_ON_ONCE(1);
return;
}
if (flags & ZPCI_PTE_INVALID) {
invalidate_pt_entry(entry);
return;
} else {
set_pt_pfaa(entry, page_addr);
validate_pt_entry(entry);
@ -146,18 +136,25 @@ static int dma_update_trans(struct zpci_dev *zdev, unsigned long pa,
u8 *page_addr = (u8 *) (pa & PAGE_MASK);
dma_addr_t start_dma_addr = dma_addr;
unsigned long irq_flags;
unsigned long *entry;
int i, rc = 0;
if (!nr_pages)
return -EINVAL;
spin_lock_irqsave(&zdev->dma_table_lock, irq_flags);
if (!zdev->dma_table)
if (!zdev->dma_table) {
rc = -EINVAL;
goto no_refresh;
}
for (i = 0; i < nr_pages; i++) {
dma_update_cpu_trans(zdev->dma_table, page_addr, dma_addr,
flags);
entry = dma_walk_cpu_trans(zdev->dma_table, dma_addr);
if (!entry) {
rc = -ENOMEM;
goto undo_cpu_trans;
}
dma_update_cpu_trans(entry, page_addr, flags);
page_addr += PAGE_SIZE;
dma_addr += PAGE_SIZE;
}
@ -176,6 +173,18 @@ static int dma_update_trans(struct zpci_dev *zdev, unsigned long pa,
rc = zpci_refresh_trans((u64) zdev->fh << 32, start_dma_addr,
nr_pages * PAGE_SIZE);
undo_cpu_trans:
if (rc && ((flags & ZPCI_PTE_VALID_MASK) == ZPCI_PTE_VALID)) {
flags = ZPCI_PTE_INVALID;
while (i-- > 0) {
page_addr -= PAGE_SIZE;
dma_addr -= PAGE_SIZE;
entry = dma_walk_cpu_trans(zdev->dma_table, dma_addr);
if (!entry)
break;
dma_update_cpu_trans(entry, page_addr, flags);
}
}
no_refresh:
spin_unlock_irqrestore(&zdev->dma_table_lock, irq_flags);
@ -260,6 +269,16 @@ out:
spin_unlock_irqrestore(&zdev->iommu_bitmap_lock, flags);
}
static inline void zpci_err_dma(unsigned long rc, unsigned long addr)
{
struct {
unsigned long rc;
unsigned long addr;
} __packed data = {rc, addr};
zpci_err_hex(&data, sizeof(data));
}
static dma_addr_t s390_dma_map_pages(struct device *dev, struct page *page,
unsigned long offset, size_t size,
enum dma_data_direction direction,
@ -270,33 +289,40 @@ static dma_addr_t s390_dma_map_pages(struct device *dev, struct page *page,
unsigned long pa = page_to_phys(page) + offset;
int flags = ZPCI_PTE_VALID;
dma_addr_t dma_addr;
int ret;
/* This rounds up number of pages based on size and offset */
nr_pages = iommu_num_pages(pa, size, PAGE_SIZE);
iommu_page_index = dma_alloc_iommu(zdev, nr_pages);
if (iommu_page_index == -1)
if (iommu_page_index == -1) {
ret = -ENOSPC;
goto out_err;
}
/* Use rounded up size */
size = nr_pages * PAGE_SIZE;
dma_addr = zdev->start_dma + iommu_page_index * PAGE_SIZE;
if (dma_addr + size > zdev->end_dma)
if (dma_addr + size > zdev->end_dma) {
ret = -ERANGE;
goto out_free;
}
if (direction == DMA_NONE || direction == DMA_TO_DEVICE)
flags |= ZPCI_TABLE_PROTECTED;
if (!dma_update_trans(zdev, pa, dma_addr, size, flags)) {
atomic64_add(nr_pages, &zdev->mapped_pages);
return dma_addr + (offset & ~PAGE_MASK);
}
ret = dma_update_trans(zdev, pa, dma_addr, size, flags);
if (ret)
goto out_free;
atomic64_add(nr_pages, &zdev->mapped_pages);
return dma_addr + (offset & ~PAGE_MASK);
out_free:
dma_free_iommu(zdev, iommu_page_index, nr_pages);
out_err:
zpci_err("map error:\n");
zpci_err_hex(&pa, sizeof(pa));
zpci_err_dma(ret, pa);
return DMA_ERROR_CODE;
}
@ -306,14 +332,16 @@ static void s390_dma_unmap_pages(struct device *dev, dma_addr_t dma_addr,
{
struct zpci_dev *zdev = to_zpci(to_pci_dev(dev));
unsigned long iommu_page_index;
int npages;
int npages, ret;
npages = iommu_num_pages(dma_addr, size, PAGE_SIZE);
dma_addr = dma_addr & PAGE_MASK;
if (dma_update_trans(zdev, 0, dma_addr, npages * PAGE_SIZE,
ZPCI_TABLE_PROTECTED | ZPCI_PTE_INVALID)) {
ret = dma_update_trans(zdev, 0, dma_addr, npages * PAGE_SIZE,
ZPCI_PTE_INVALID);
if (ret) {
zpci_err("unmap error:\n");
zpci_err_hex(&dma_addr, sizeof(dma_addr));
zpci_err_dma(ret, dma_addr);
return;
}
atomic64_add(npages, &zdev->unmapped_pages);

3
arch/x86/include/asm/msr-index.h
View File

@ -35,7 +35,7 @@
#define MSR_IA32_PERFCTR0 0x000000c1
#define MSR_IA32_PERFCTR1 0x000000c2
#define MSR_FSB_FREQ 0x000000cd
#define MSR_NHM_PLATFORM_INFO 0x000000ce
#define MSR_PLATFORM_INFO 0x000000ce
#define MSR_NHM_SNB_PKG_CST_CFG_CTL 0x000000e2
#define NHM_C3_AUTO_DEMOTE (1UL << 25)
@ -44,7 +44,6 @@
#define SNB_C1_AUTO_UNDEMOTE (1UL << 27)
#define SNB_C3_AUTO_UNDEMOTE (1UL << 28)
#define MSR_PLATFORM_INFO 0x000000ce
#define MSR_MTRRcap 0x000000fe
#define MSR_IA32_BBL_CR_CTL 0x00000119
#define MSR_IA32_BBL_CR_CTL3 0x0000011e

3
arch/x86/kernel/cpu/common.c
View File

@ -273,10 +273,9 @@ __setup("nosmap", setup_disable_smap);
static __always_inline void setup_smap(struct cpuinfo_x86 *c)
{
unsigned long eflags;
unsigned long eflags = native_save_fl();
/* This should have been cleared long ago */
raw_local_save_flags(eflags);
BUG_ON(eflags & X86_EFLAGS_AC);
if (cpu_has(c, X86_FEATURE_SMAP)) {

11
arch/x86/kernel/fpu/signal.c
View File

@ -385,20 +385,19 @@ fpu__alloc_mathframe(unsigned long sp, int ia32_frame,
*/
void fpu__init_prepare_fx_sw_frame(void)
{
int fsave_header_size = sizeof(struct fregs_state);
int size = xstate_size + FP_XSTATE_MAGIC2_SIZE;
if (config_enabled(CONFIG_X86_32))
size += fsave_header_size;
fx_sw_reserved.magic1 = FP_XSTATE_MAGIC1;
fx_sw_reserved.extended_size = size;
fx_sw_reserved.xfeatures = xfeatures_mask;
fx_sw_reserved.xstate_size = xstate_size;
if (config_enabled(CONFIG_IA32_EMULATION)) {
if (config_enabled(CONFIG_IA32_EMULATION) ||
config_enabled(CONFIG_X86_32)) {
int fsave_header_size = sizeof(struct fregs_state);
fx_sw_reserved_ia32 = fx_sw_reserved;
fx_sw_reserved_ia32.extended_size += fsave_header_size;
fx_sw_reserved_ia32.extended_size = size + fsave_header_size;
}
}

1
arch/x86/kernel/fpu/xstate.c
View File

@ -694,7 +694,6 @@ void *get_xsave_addr(struct xregs_state *xsave, int xstate_feature)
if (!boot_cpu_has(X86_FEATURE_XSAVE))
return NULL;
xsave = &current->thread.fpu.state.xsave;
/*
* We should not ever be requesting features that we
* have not enabled. Remember that pcntxt_mask is

6
arch/x86/kernel/mcount_64.S
View File

@ -278,6 +278,12 @@ trace:
/* save_mcount_regs fills in first two parameters */
save_mcount_regs
/*
* When DYNAMIC_FTRACE is not defined, ARCH_SUPPORTS_FTRACE_OPS is not
* set (see include/asm/ftrace.h and include/linux/ftrace.h). Only the
* ip and parent ip are used and the list function is called when
* function tracing is enabled.
*/
call *ftrace_trace_function
restore_mcount_regs

47
arch/x86/mm/mpx.c
View File

@ -585,6 +585,29 @@ static unsigned long mpx_bd_entry_to_bt_addr(struct mm_struct *mm,
return bt_addr;
}
/*
* We only want to do a 4-byte get_user() on 32-bit. Otherwise,
* we might run off the end of the bounds table if we are on
* a 64-bit kernel and try to get 8 bytes.
*/
int get_user_bd_entry(struct mm_struct *mm, unsigned long *bd_entry_ret,
long __user *bd_entry_ptr)
{
u32 bd_entry_32;
int ret;
if (is_64bit_mm(mm))
return get_user(*bd_entry_ret, bd_entry_ptr);
/*
* Note that get_user() uses the type of the *pointer* to
* establish the size of the get, not the destination.
*/
ret = get_user(bd_entry_32, (u32 __user *)bd_entry_ptr);
*bd_entry_ret = bd_entry_32;
return ret;
}
/*
* Get the base of bounds tables pointed by specific bounds
* directory entry.
@ -605,7 +628,7 @@ static int get_bt_addr(struct mm_struct *mm,
int need_write = 0;
pagefault_disable();
ret = get_user(bd_entry, bd_entry_ptr);
ret = get_user_bd_entry(mm, &bd_entry, bd_entry_ptr);
pagefault_enable();
if (!ret)
break;
@ -700,11 +723,23 @@ static unsigned long mpx_get_bt_entry_offset_bytes(struct mm_struct *mm,
*/
static inline unsigned long bd_entry_virt_space(struct mm_struct *mm)
{
unsigned long long virt_space = (1ULL << boot_cpu_data.x86_virt_bits);
if (is_64bit_mm(mm))
return virt_space / MPX_BD_NR_ENTRIES_64;
else
return virt_space / MPX_BD_NR_ENTRIES_32;
unsigned long long virt_space;
unsigned long long GB = (1ULL << 30);
/*
* This covers 32-bit emulation as well as 32-bit kernels
* running on 64-bit harware.
*/
if (!is_64bit_mm(mm))
return (4ULL * GB) / MPX_BD_NR_ENTRIES_32;
/*
* 'x86_virt_bits' returns what the hardware is capable
* of, and returns the full >32-bit adddress space when
* running 32-bit kernels on 64-bit hardware.
*/
virt_space = (1ULL << boot_cpu_data.x86_virt_bits);
return virt_space / MPX_BD_NR_ENTRIES_64;
}
/*

2
block/blk.h
View File

@ -72,8 +72,6 @@ void blk_dequeue_request(struct request *rq);
void __blk_queue_free_tags(struct request_queue *q);
bool __blk_end_bidi_request(struct request *rq, int error,
unsigned int nr_bytes, unsigned int bidi_bytes);
int blk_queue_enter(struct request_queue *q, gfp_t gfp);
void blk_queue_exit(struct request_queue *q);
void blk_freeze_queue(struct request_queue *q);
static inline void blk_queue_enter_live(struct request_queue *q)

2
drivers/acpi/cppc_acpi.c
View File

@ -304,7 +304,7 @@ EXPORT_SYMBOL_GPL(acpi_get_psd_map);
static int register_pcc_channel(int pcc_subspace_idx)
{
struct acpi_pcct_subspace *cppc_ss;
struct acpi_pcct_hw_reduced *cppc_ss;
unsigned int len;
if (pcc_subspace_idx >= 0) {

2
drivers/acpi/ec.c
View File

@ -1103,7 +1103,7 @@ static int acpi_ec_query(struct acpi_ec *ec, u8 *data)
}
err_exit:
if (result && q)
if (result)
acpi_ec_delete_query(q);
if (data)
*data = value;

48
drivers/acpi/sbshc.c
View File

@ -14,7 +14,6 @@
#include <linux/delay.h>
#include <linux/module.h>
#include <linux/interrupt.h>
#include <linux/dmi.h>
#include "sbshc.h"
#define PREFIX "ACPI: "
@ -30,6 +29,7 @@ struct acpi_smb_hc {
u8 query_bit;
smbus_alarm_callback callback;
void *context;
bool done;
};
static int acpi_smbus_hc_add(struct acpi_device *device);
@ -88,8 +88,6 @@ enum acpi_smb_offset {
ACPI_SMB_ALARM_DATA = 0x26, /* 2 bytes alarm data */
};
static bool macbook;
static inline int smb_hc_read(struct acpi_smb_hc *hc, u8 address, u8 *data)
{
return ec_read(hc->offset + address, data);
@ -100,27 +98,11 @@ static inline int smb_hc_write(struct acpi_smb_hc *hc, u8 address, u8 data)
return ec_write(hc->offset + address, data);
}
static inline int smb_check_done(struct acpi_smb_hc *hc)
{
union acpi_smb_status status = {.raw = 0};
smb_hc_read(hc, ACPI_SMB_STATUS, &status.raw);
return status.fields.done && (status.fields.status == SMBUS_OK);
}
static int wait_transaction_complete(struct acpi_smb_hc *hc, int timeout)
{
if (wait_event_timeout(hc->wait, smb_check_done(hc),
msecs_to_jiffies(timeout)))
if (wait_event_timeout(hc->wait, hc->done, msecs_to_jiffies(timeout)))
return 0;
/*
* After the timeout happens, OS will try to check the status of SMbus.
* If the status is what OS expected, it will be regarded as the bogus
* timeout.
*/
if (smb_check_done(hc))
return 0;
else
return -ETIME;
return -ETIME;
}
static int acpi_smbus_transaction(struct acpi_smb_hc *hc, u8 protocol,
@ -135,8 +117,7 @@ static int acpi_smbus_transaction(struct acpi_smb_hc *hc, u8 protocol,
}
mutex_lock(&hc->lock);
if (macbook)
udelay(5);
hc->done = false;
if (smb_hc_read(hc, ACPI_SMB_PROTOCOL, &temp))
goto end;
if (temp) {
@ -235,8 +216,10 @@ static int smbus_alarm(void *context)
if (smb_hc_read(hc, ACPI_SMB_STATUS, &status.raw))
return 0;
/* Check if it is only a completion notify */
if (status.fields.done)
if (status.fields.done && status.fields.status == SMBUS_OK) {
hc->done = true;
wake_up(&hc->wait);
}
if (!status.fields.alarm)
return 0;
mutex_lock(&hc->lock);
@ -262,29 +245,12 @@ extern int acpi_ec_add_query_handler(struct acpi_ec *ec, u8 query_bit,
acpi_handle handle, acpi_ec_query_func func,
void *data);
static int macbook_dmi_match(const struct dmi_system_id *d)
{
pr_debug("Detected MacBook, enabling workaround\n");
macbook = true;
return 0;
}
static struct dmi_system_id acpi_smbus_dmi_table[] = {
{ macbook_dmi_match, "Apple MacBook", {
DMI_MATCH(DMI_BOARD_VENDOR, "Apple"),
DMI_MATCH(DMI_PRODUCT_NAME, "MacBook") },
},
{ },
};
static int acpi_smbus_hc_add(struct acpi_device *device)
{
int status;
unsigned long long val;
struct acpi_smb_hc *hc;
dmi_check_system(acpi_smbus_dmi_table);
if (!device)
return -EINVAL;

6
drivers/base/power/wakeirq.c
View File

@ -68,6 +68,9 @@ int dev_pm_set_wake_irq(struct device *dev, int irq)
struct wake_irq *wirq;
int err;
if (irq < 0)
return -EINVAL;
wirq = kzalloc(sizeof(*wirq), GFP_KERNEL);
if (!wirq)
return -ENOMEM;
@ -167,6 +170,9 @@ int dev_pm_set_dedicated_wake_irq(struct device *dev, int irq)
struct wake_irq *wirq;
int err;
if (irq < 0)
return -EINVAL;
wirq = kzalloc(sizeof(*wirq), GFP_KERNEL);
if (!wirq)
return -ENOMEM;

82
drivers/char/ipmi/ipmi_si_intf.c
View File

@ -412,18 +412,42 @@ static enum si_sm_result start_next_msg(struct smi_info *smi_info)
return rv;
}
static void start_check_enables(struct smi_info *smi_info)
static void smi_mod_timer(struct smi_info *smi_info, unsigned long new_val)
{
smi_info->last_timeout_jiffies = jiffies;
mod_timer(&smi_info->si_timer, new_val);
smi_info->timer_running = true;
}
/*
* Start a new message and (re)start the timer and thread.
*/
static void start_new_msg(struct smi_info *smi_info, unsigned char *msg,
unsigned int size)
{
smi_mod_timer(smi_info, jiffies + SI_TIMEOUT_JIFFIES);
if (smi_info->thread)
wake_up_process(smi_info->thread);
smi_info->handlers->start_transaction(smi_info->si_sm, msg, size);
}
static void start_check_enables(struct smi_info *smi_info, bool start_timer)
{
unsigned char msg[2];
msg[0] = (IPMI_NETFN_APP_REQUEST << 2);
msg[1] = IPMI_GET_BMC_GLOBAL_ENABLES_CMD;
smi_info->handlers->start_transaction(smi_info->si_sm, msg, 2);
if (start_timer)
start_new_msg(smi_info, msg, 2);
else
smi_info->handlers->start_transaction(smi_info->si_sm, msg, 2);
smi_info->si_state = SI_CHECKING_ENABLES;
}
static void start_clear_flags(struct smi_info *smi_info)
static void start_clear_flags(struct smi_info *smi_info, bool start_timer)
{
unsigned char msg[3];
@ -432,7 +456,10 @@ static void start_clear_flags(struct smi_info *smi_info)
msg[1] = IPMI_CLEAR_MSG_FLAGS_CMD;
msg[2] = WDT_PRE_TIMEOUT_INT;
smi_info->handlers->start_transaction(smi_info->si_sm, msg, 3);
if (start_timer)
start_new_msg(smi_info, msg, 3);
else
smi_info->handlers->start_transaction(smi_info->si_sm, msg, 3);
smi_info->si_state = SI_CLEARING_FLAGS;
}
@ -442,10 +469,8 @@ static void start_getting_msg_queue(struct smi_info *smi_info)
smi_info->curr_msg->data[1] = IPMI_GET_MSG_CMD;
smi_info->curr_msg->data_size = 2;
smi_info->handlers->start_transaction(
smi_info->si_sm,
smi_info->curr_msg->data,
smi_info->curr_msg->data_size);
start_new_msg(smi_info, smi_info->curr_msg->data,
smi_info->curr_msg->data_size);
smi_info->si_state = SI_GETTING_MESSAGES;
}
@ -455,20 +480,11 @@ static void start_getting_events(struct smi_info *smi_info)
smi_info->curr_msg->data[1] = IPMI_READ_EVENT_MSG_BUFFER_CMD;
smi_info->curr_msg->data_size = 2;
smi_info->handlers->start_transaction(
smi_info->si_sm,
smi_info->curr_msg->data,
smi_info->curr_msg->data_size);
start_new_msg(smi_info, smi_info->curr_msg->data,
smi_info->curr_msg->data_size);
smi_info->si_state = SI_GETTING_EVENTS;
}
static void smi_mod_timer(struct smi_info *smi_info, unsigned long new_val)
{
smi_info->last_timeout_jiffies = jiffies;
mod_timer(&smi_info->si_timer, new_val);
smi_info->timer_running = true;
}
/*
* When we have a situtaion where we run out of memory and cannot
* allocate messages, we just leave them in the BMC and run the system
@ -478,11 +494,11 @@ static void smi_mod_timer(struct smi_info *smi_info, unsigned long new_val)
* Note that we cannot just use disable_irq(), since the interrupt may
* be shared.
*/
static inline bool disable_si_irq(struct smi_info *smi_info)
static inline bool disable_si_irq(struct smi_info *smi_info, bool start_timer)
{
if ((smi_info->irq) && (!smi_info->interrupt_disabled)) {
smi_info->interrupt_disabled = true;
start_check_enables(smi_info);
start_check_enables(smi_info, start_timer);
return true;
}
return false;
@ -492,7 +508,7 @@ static inline bool enable_si_irq(struct smi_info *smi_info)
{
if ((smi_info->irq) && (smi_info->interrupt_disabled)) {
smi_info->interrupt_disabled = false;
start_check_enables(smi_info);
start_check_enables(smi_info, true);
return true;
}
return false;
@ -510,7 +526,7 @@ static struct ipmi_smi_msg *alloc_msg_handle_irq(struct smi_info *smi_info)
msg = ipmi_alloc_smi_msg();
if (!msg) {
if (!disable_si_irq(smi_info))
if (!disable_si_irq(smi_info, true))
smi_info->si_state = SI_NORMAL;
} else if (enable_si_irq(smi_info)) {
ipmi_free_smi_msg(msg);
@ -526,7 +542,7 @@ static void handle_flags(struct smi_info *smi_info)
/* Watchdog pre-timeout */
smi_inc_stat(smi_info, watchdog_pretimeouts);
start_clear_flags(smi_info);
start_clear_flags(smi_info, true);
smi_info->msg_flags &= ~WDT_PRE_TIMEOUT_INT;
if (smi_info->intf)
ipmi_smi_watchdog_pretimeout(smi_info->intf);
@ -879,8 +895,7 @@ static enum si_sm_result smi_event_handler(struct smi_info *smi_info,
msg[0] = (IPMI_NETFN_APP_REQUEST << 2);
msg[1] = IPMI_GET_MSG_FLAGS_CMD;
smi_info->handlers->start_transaction(
smi_info->si_sm, msg, 2);
start_new_msg(smi_info, msg, 2);
smi_info->si_state = SI_GETTING_FLAGS;
goto restart;
}
@ -910,7 +925,7 @@ static enum si_sm_result smi_event_handler(struct smi_info *smi_info,
* disable and messages disabled.
*/
if (smi_info->supports_event_msg_buff || smi_info->irq) {
start_check_enables(smi_info);
start_check_enables(smi_info, true);
} else {
smi_info->curr_msg = alloc_msg_handle_irq(smi_info);
if (!smi_info->curr_msg)
@ -920,6 +935,13 @@ static enum si_sm_result smi_event_handler(struct smi_info *smi_info,
}
goto restart;
}
if (si_sm_result == SI_SM_IDLE && smi_info->timer_running) {
/* Ok it if fails, the timer will just go off. */
if (del_timer(&smi_info->si_timer))
smi_info->timer_running = false;
}
out:
return si_sm_result;
}
@ -2560,6 +2582,7 @@ static const struct of_device_id of_ipmi_match[] = {
.data = (void *)(unsigned long) SI_BT },
{},
};
MODULE_DEVICE_TABLE(of, of_ipmi_match);
static int of_ipmi_probe(struct platform_device *dev)
{
@ -2646,7 +2669,6 @@ static int of_ipmi_probe(struct platform_device *dev)
}
return 0;
}
MODULE_DEVICE_TABLE(of, of_ipmi_match);
#else
#define of_ipmi_match NULL
static int of_ipmi_probe(struct platform_device *dev)
@ -3613,7 +3635,7 @@ static int try_smi_init(struct smi_info *new_smi)
* Start clearing the flags before we enable interrupts or the
* timer to avoid racing with the timer.
*/
start_clear_flags(new_smi);
start_clear_flags(new_smi, false);
/*
* IRQ is defined to be set when non-zero. req_events will
@ -3908,7 +3930,7 @@ static void cleanup_one_si(struct smi_info *to_clean)
poll(to_clean);
schedule_timeout_uninterruptible(1);
}
disable_si_irq(to_clean);
disable_si_irq(to_clean, false);
while (to_clean->curr_msg || (to_clean->si_state != SI_NORMAL)) {
poll(to_clean);
schedule_timeout_uninterruptible(1);

8
drivers/char/ipmi/ipmi_watchdog.c
View File

@ -153,6 +153,9 @@ static int timeout = 10;
/* The pre-timeout is disabled by default. */
static int pretimeout;
/* Default timeout to set on panic */
static int panic_wdt_timeout = 255;
/* Default action is to reset the board on a timeout. */
static unsigned char action_val = WDOG_TIMEOUT_RESET;
@ -293,6 +296,9 @@ MODULE_PARM_DESC(timeout, "Timeout value in seconds.");
module_param(pretimeout, timeout, 0644);
MODULE_PARM_DESC(pretimeout, "Pretimeout value in seconds.");
module_param(panic_wdt_timeout, timeout, 0644);
MODULE_PARM_DESC(timeout, "Timeout value on kernel panic in seconds.");
module_param_cb(action, &param_ops_str, action_op, 0644);
MODULE_PARM_DESC(action, "Timeout action. One of: "
"reset, none, power_cycle, power_off.");
@ -1189,7 +1195,7 @@ static int wdog_panic_handler(struct notifier_block *this,
/* Make sure we do this only once. */
panic_event_handled = 1;
timeout = 255;
timeout = panic_wdt_timeout;
pretimeout = 0;
panic_halt_ipmi_set_timeout();
}

1
drivers/clocksource/Kconfig
View File

@ -1,4 +1,5 @@
menu "Clock Source drivers"
depends on !ARCH_USES_GETTIMEOFFSET
config CLKSRC_OF
bool

4
drivers/clocksource/fsl_ftm_timer.c
View File

@ -203,7 +203,7 @@ static int __init ftm_clockevent_init(unsigned long freq, int irq)
int err;
ftm_writel(0x00, priv->clkevt_base + FTM_CNTIN);
ftm_writel(~0UL, priv->clkevt_base + FTM_MOD);
ftm_writel(~0u, priv->clkevt_base + FTM_MOD);
ftm_reset_counter(priv->clkevt_base);
@ -230,7 +230,7 @@ static int __init ftm_clocksource_init(unsigned long freq)
int err;
ftm_writel(0x00, priv->clksrc_base + FTM_CNTIN);
ftm_writel(~0UL, priv->clksrc_base + FTM_MOD);
ftm_writel(~0u, priv->clksrc_base + FTM_MOD);
ftm_reset_counter(priv->clksrc_base);

1
drivers/cpufreq/Kconfig.arm
View File

@ -84,6 +84,7 @@ config ARM_KIRKWOOD_CPUFREQ
config ARM_MT8173_CPUFREQ
bool "Mediatek MT8173 CPUFreq support"
depends on ARCH_MEDIATEK && REGULATOR
depends on ARM64 || (ARM_CPU_TOPOLOGY && COMPILE_TEST)
depends on !CPU_THERMAL || THERMAL=y
select PM_OPP
help

1
drivers/cpufreq/Kconfig.x86
View File

@ -5,7 +5,6 @@
config X86_INTEL_PSTATE
bool "Intel P state control"
depends on X86
select ACPI_PROCESSOR if ACPI
help
This driver provides a P state for Intel core processors.
The driver implements an internal governor and will become

318
drivers/cpufreq/intel_pstate.c
View File

@ -34,14 +34,10 @@
#include <asm/cpu_device_id.h>
#include <asm/cpufeature.h>
#if IS_ENABLED(CONFIG_ACPI)
#include <acpi/processor.h>
#endif
#define BYT_RATIOS 0x66a
#define BYT_VIDS 0x66b
#define BYT_TURBO_RATIOS 0x66c
#define BYT_TURBO_VIDS 0x66d
#define ATOM_RATIOS 0x66a
#define ATOM_VIDS 0x66b
#define ATOM_TURBO_RATIOS 0x66c
#define ATOM_TURBO_VIDS 0x66d
#define FRAC_BITS 8
#define int_tofp(X) ((int64_t)(X) << FRAC_BITS)
@ -117,9 +113,6 @@ struct cpudata {
u64 prev_mperf;
u64 prev_tsc;
struct sample sample;
#if IS_ENABLED(CONFIG_ACPI)
struct acpi_processor_performance acpi_perf_data;
#endif
};
static struct cpudata **all_cpu_data;
@ -150,7 +143,6 @@ struct cpu_defaults {
static struct pstate_adjust_policy pid_params;
static struct pstate_funcs pstate_funcs;
static int hwp_active;
static int no_acpi_perf;
struct perf_limits {
int no_turbo;
@ -163,8 +155,6 @@ struct perf_limits {
int max_sysfs_pct;
int min_policy_pct;
int min_sysfs_pct;
int max_perf_ctl;
int min_perf_ctl;
};
static struct perf_limits performance_limits = {
@ -191,8 +181,6 @@ static struct perf_limits powersave_limits = {
.max_sysfs_pct = 100,
.min_policy_pct = 0,
.min_sysfs_pct = 0,
.max_perf_ctl = 0,
.min_perf_ctl = 0,
};
#ifdef CONFIG_CPU_FREQ_DEFAULT_GOV_PERFORMANCE
@ -201,153 +189,6 @@ static struct perf_limits *limits = &performance_limits;
static struct perf_limits *limits = &powersave_limits;
#endif
#if IS_ENABLED(CONFIG_ACPI)
/*
* The max target pstate ratio is a 8 bit value in both PLATFORM_INFO MSR and
* in TURBO_RATIO_LIMIT MSR, which pstate driver stores in max_pstate and
* max_turbo_pstate fields. The PERF_CTL MSR contains 16 bit value for P state
* ratio, out of it only high 8 bits are used. For example 0x1700 is setting
* target ratio 0x17. The _PSS control value stores in a format which can be
* directly written to PERF_CTL MSR. But in intel_pstate driver this shift
* occurs during write to PERF_CTL (E.g. for cores core_set_pstate()).
* This function converts the _PSS control value to intel pstate driver format
* for comparison and assignment.
*/
static int convert_to_native_pstate_format(struct cpudata *cpu, int index)
{
return cpu->acpi_perf_data.states[index].control >> 8;
}
static int intel_pstate_init_perf_limits(struct cpufreq_policy *policy)
{
struct cpudata *cpu;
int ret;
bool turbo_absent = false;
int max_pstate_index;
int min_pss_ctl, max_pss_ctl, turbo_pss_ctl;
int i;
cpu = all_cpu_data[policy->cpu];
pr_debug("intel_pstate: default limits 0x%x 0x%x 0x%x\n",
cpu->pstate.min_pstate, cpu->pstate.max_pstate,
cpu->pstate.turbo_pstate);
if (!cpu->acpi_perf_data.shared_cpu_map &&
zalloc_cpumask_var_node(&cpu->acpi_perf_data.shared_cpu_map,
GFP_KERNEL, cpu_to_node(policy->cpu))) {
return -ENOMEM;
}
ret = acpi_processor_register_performance(&cpu->acpi_perf_data,
policy->cpu);
if (ret)
return ret;
/*
* Check if the control value in _PSS is for PERF_CTL MSR, which should
* guarantee that the states returned by it map to the states in our
* list directly.
*/
if (cpu->acpi_perf_data.control_register.space_id !=
ACPI_ADR_SPACE_FIXED_HARDWARE)
return -EIO;
pr_debug("intel_pstate: CPU%u - ACPI _PSS perf data\n", policy->cpu);
for (i = 0; i < cpu->acpi_perf_data.state_count; i++)
pr_debug(" %cP%d: %u MHz, %u mW, 0x%x\n",
(i == cpu->acpi_perf_data.state ? '*' : ' '), i,
(u32) cpu->acpi_perf_data.states[i].core_frequency,
(u32) cpu->acpi_perf_data.states[i].power,
(u32) cpu->acpi_perf_data.states[i].control);
/*
* If there is only one entry _PSS, simply ignore _PSS and continue as
* usual without taking _PSS into account
*/
if (cpu->acpi_perf_data.state_count < 2)
return 0;
turbo_pss_ctl = convert_to_native_pstate_format(cpu, 0);
min_pss_ctl = convert_to_native_pstate_format(cpu,
cpu->acpi_perf_data.state_count - 1);
/* Check if there is a turbo freq in _PSS */
if (turbo_pss_ctl <= cpu->pstate.max_pstate &&
turbo_pss_ctl > cpu->pstate.min_pstate) {
pr_debug("intel_pstate: no turbo range exists in _PSS\n");
limits->no_turbo = limits->turbo_disabled = 1;
cpu->pstate.turbo_pstate = cpu->pstate.max_pstate;
turbo_absent = true;
}
/* Check if the max non turbo p state < Intel P state max */
max_pstate_index = turbo_absent ? 0 : 1;
max_pss_ctl = convert_to_native_pstate_format(cpu, max_pstate_index);
if (max_pss_ctl < cpu->pstate.max_pstate &&
max_pss_ctl > cpu->pstate.min_pstate)
cpu->pstate.max_pstate = max_pss_ctl;
/* check If min perf > Intel P State min */
if (min_pss_ctl > cpu->pstate.min_pstate &&
min_pss_ctl < cpu->pstate.max_pstate) {
cpu->pstate.min_pstate = min_pss_ctl;
policy->cpuinfo.min_freq = min_pss_ctl * cpu->pstate.scaling;
}
if (turbo_absent)
policy->cpuinfo.max_freq = cpu->pstate.max_pstate *
cpu->pstate.scaling;
else {
policy->cpuinfo.max_freq = cpu->pstate.turbo_pstate *
cpu->pstate.scaling;
/*
* The _PSS table doesn't contain whole turbo frequency range.
* This just contains +1 MHZ above the max non turbo frequency,
* with control value corresponding to max turbo ratio. But
* when cpufreq set policy is called, it will call with this
* max frequency, which will cause a reduced performance as
* this driver uses real max turbo frequency as the max
* frequeny. So correct this frequency in _PSS table to
* correct max turbo frequency based on the turbo ratio.
* Also need to convert to MHz as _PSS freq is in MHz.
*/
cpu->acpi_perf_data.states[0].core_frequency =
turbo_pss_ctl * 100;
}
pr_debug("intel_pstate: Updated limits using _PSS 0x%x 0x%x 0x%x\n",
cpu->pstate.min_pstate, cpu->pstate.max_pstate,
cpu->pstate.turbo_pstate);
pr_debug("intel_pstate: policy max_freq=%d Khz min_freq = %d KHz\n",
policy->cpuinfo.max_freq, policy->cpuinfo.min_freq);
return 0;
}
static int intel_pstate_exit_perf_limits(struct cpufreq_policy *policy)
{
struct cpudata *cpu;
if (!no_acpi_perf)
return 0;
cpu = all_cpu_data[policy->cpu];
acpi_processor_unregister_performance(policy->cpu);
return 0;
}
#else
static int intel_pstate_init_perf_limits(struct cpufreq_policy *policy)
{
return 0;
}
static int intel_pstate_exit_perf_limits(struct cpufreq_policy *policy)
{
return 0;
}
#endif
static inline void pid_reset(struct _pid *pid, int setpoint, int busy,
int deadband, int integral) {
pid->setpoint = setpoint;
@ -687,31 +528,31 @@ static void intel_pstate_hwp_enable(struct cpudata *cpudata)
wrmsrl_on_cpu(cpudata->cpu, MSR_PM_ENABLE, 0x1);
}
static int byt_get_min_pstate(void)
static int atom_get_min_pstate(void)
{
u64 value;
rdmsrl(BYT_RATIOS, value);
rdmsrl(ATOM_RATIOS, value);
return (value >> 8) & 0x7F;
}
static int byt_get_max_pstate(void)
static int atom_get_max_pstate(void)
{
u64 value;
rdmsrl(BYT_RATIOS, value);
rdmsrl(ATOM_RATIOS, value);
return (value >> 16) & 0x7F;
}
static int byt_get_turbo_pstate(void)
static int atom_get_turbo_pstate(void)
{
u64 value;
rdmsrl(BYT_TURBO_RATIOS, value);
rdmsrl(ATOM_TURBO_RATIOS, value);
return value & 0x7F;
}
static void byt_set_pstate(struct cpudata *cpudata, int pstate)
static void atom_set_pstate(struct cpudata *cpudata, int pstate)
{
u64 val;
int32_t vid_fp;
@ -736,27 +577,42 @@ static void byt_set_pstate(struct cpudata *cpudata, int pstate)
wrmsrl_on_cpu(cpudata->cpu, MSR_IA32_PERF_CTL, val);
}
#define BYT_BCLK_FREQS 5
static int byt_freq_table[BYT_BCLK_FREQS] = { 833, 1000, 1333, 1167, 800};
static int byt_get_scaling(void)
static int silvermont_get_scaling(void)
{
u64 value;
int i;
/* Defined in Table 35-6 from SDM (Sept 2015) */
static int silvermont_freq_table[] = {
83300, 100000, 133300, 116700, 80000};
rdmsrl(MSR_FSB_FREQ, value);
i = value & 0x3;
i = value & 0x7;
WARN_ON(i > 4);
BUG_ON(i > BYT_BCLK_FREQS);
return byt_freq_table[i] * 100;
return silvermont_freq_table[i];
}
static void byt_get_vid(struct cpudata *cpudata)
static int airmont_get_scaling(void)
{
u64 value;
int i;
/* Defined in Table 35-10 from SDM (Sept 2015) */
static int airmont_freq_table[] = {
83300, 100000, 133300, 116700, 80000,
93300, 90000, 88900, 87500};
rdmsrl(MSR_FSB_FREQ, value);
i = value & 0xF;
WARN_ON(i > 8);
return airmont_freq_table[i];
}
static void atom_get_vid(struct cpudata *cpudata)
{
u64 value;
rdmsrl(BYT_VIDS, value);
rdmsrl(ATOM_VIDS, value);
cpudata->vid.min = int_tofp((value >> 8) & 0x7f);
cpudata->vid.max = int_tofp((value >> 16) & 0x7f);
cpudata->vid.ratio = div_fp(
@ -764,7 +620,7 @@ static void byt_get_vid(struct cpudata *cpudata)
int_tofp(cpudata->pstate.max_pstate -
cpudata->pstate.min_pstate));
rdmsrl(BYT_TURBO_VIDS, value);
rdmsrl(ATOM_TURBO_VIDS, value);
cpudata->vid.turbo = value & 0x7f;
}
@ -885,7 +741,7 @@ static struct cpu_defaults core_params = {
},
};
static struct cpu_defaults byt_params = {
static struct cpu_defaults silvermont_params = {
.pid_policy = {
.sample_rate_ms = 10,
.deadband = 0,
@ -895,13 +751,33 @@ static struct cpu_defaults byt_params = {
.i_gain_pct = 4,
},
.funcs = {
.get_max = byt_get_max_pstate,
.get_max_physical = byt_get_max_pstate,
.get_min = byt_get_min_pstate,
.get_turbo = byt_get_turbo_pstate,
.set = byt_set_pstate,
.get_scaling = byt_get_scaling,
.get_vid = byt_get_vid,
.get_max = atom_get_max_pstate,
.get_max_physical = atom_get_max_pstate,
.get_min = atom_get_min_pstate,
.get_turbo = atom_get_turbo_pstate,
.set = atom_set_pstate,
.get_scaling = silvermont_get_scaling,
.get_vid = atom_get_vid,
},
};
static struct cpu_defaults airmont_params = {
.pid_policy = {
.sample_rate_ms = 10,
.deadband = 0,
.setpoint = 60,
.p_gain_pct = 14,
.d_gain_pct = 0,
.i_gain_pct = 4,
},
.funcs = {
.get_max = atom_get_max_pstate,
.get_max_physical = atom_get_max_pstate,
.get_min = atom_get_min_pstate,
.get_turbo = atom_get_turbo_pstate,
.set = atom_set_pstate,
.get_scaling = airmont_get_scaling,
.get_vid = atom_get_vid,
},
};
@ -938,23 +814,12 @@ static void intel_pstate_get_min_max(struct cpudata *cpu, int *min, int *max)
* policy, or by cpu specific default values determined through
* experimentation.
*/
if (limits->max_perf_ctl && limits->max_sysfs_pct >=
limits->max_policy_pct) {
*max = limits->max_perf_ctl;
} else {
max_perf_adj = fp_toint(mul_fp(int_tofp(max_perf),
limits->max_perf));
*max = clamp_t(int, max_perf_adj, cpu->pstate.min_pstate,
cpu->pstate.turbo_pstate);
}
max_perf_adj = fp_toint(mul_fp(int_tofp(max_perf), limits->max_perf));
*max = clamp_t(int, max_perf_adj,
cpu->pstate.min_pstate, cpu->pstate.turbo_pstate);
if (limits->min_perf_ctl) {
*min = limits->min_perf_ctl;
} else {
min_perf = fp_toint(mul_fp(int_tofp(max_perf),
limits->min_perf));
*min = clamp_t(int, min_perf, cpu->pstate.min_pstate, max_perf);
}
min_perf = fp_toint(mul_fp(int_tofp(max_perf), limits->min_perf));
*min = clamp_t(int, min_perf, cpu->pstate.min_pstate, max_perf);
}
static void intel_pstate_set_pstate(struct cpudata *cpu, int pstate, bool force)
@ -1153,7 +1018,7 @@ static void intel_pstate_timer_func(unsigned long __data)
static const struct x86_cpu_id intel_pstate_cpu_ids[] = {
ICPU(0x2a, core_params),
ICPU(0x2d, core_params),
ICPU(0x37, byt_params),
ICPU(0x37, silvermont_params),
ICPU(0x3a, core_params),
ICPU(0x3c, core_params),
ICPU(0x3d, core_params),
@ -1162,7 +1027,7 @@ static const struct x86_cpu_id intel_pstate_cpu_ids[] = {
ICPU(0x45, core_params),
ICPU(0x46, core_params),
ICPU(0x47, core_params),
ICPU(0x4c, byt_params),
ICPU(0x4c, airmont_params),
ICPU(0x4e, core_params),
ICPU(0x4f, core_params),
ICPU(0x5e, core_params),
@ -1229,12 +1094,6 @@ static unsigned int intel_pstate_get(unsigned int cpu_num)
static int intel_pstate_set_policy(struct cpufreq_policy *policy)
{
#if IS_ENABLED(CONFIG_ACPI)
struct cpudata *cpu;
int i;
#endif
pr_debug("intel_pstate: %s max %u policy->max %u\n", __func__,
policy->cpuinfo.max_freq, policy->max);
if (!policy->cpuinfo.max_freq)
return -ENODEV;
@ -1270,23 +1129,6 @@ static int intel_pstate_set_policy(struct cpufreq_policy *policy)
limits->max_perf = div_fp(int_tofp(limits->max_perf_pct),
int_tofp(100));
#if IS_ENABLED(CONFIG_ACPI)
cpu = all_cpu_data[policy->cpu];
for (i = 0; i < cpu->acpi_perf_data.state_count; i++) {
int control;
control = convert_to_native_pstate_format(cpu, i);
if (control * cpu->pstate.scaling == policy->max)
limits->max_perf_ctl = control;
if (control * cpu->pstate.scaling == policy->min)
limits->min_perf_ctl = control;
}
pr_debug("intel_pstate: max %u policy_max %u perf_ctl [0x%x-0x%x]\n",
policy->cpuinfo.max_freq, policy->max, limits->min_perf_ctl,
limits->max_perf_ctl);
#endif
if (hwp_active)
intel_pstate_hwp_set();
@ -1341,30 +1183,18 @@ static int intel_pstate_cpu_init(struct cpufreq_policy *policy)
policy->cpuinfo.min_freq = cpu->pstate.min_pstate * cpu->pstate.scaling;
policy->cpuinfo.max_freq =
cpu->pstate.turbo_pstate * cpu->pstate.scaling;
if (!no_acpi_perf)
intel_pstate_init_perf_limits(policy);
/*
* If there is no acpi perf data or error, we ignore and use Intel P
* state calculated limits, So this is not fatal error.
*/
policy->cpuinfo.transition_latency = CPUFREQ_ETERNAL;
cpumask_set_cpu(policy->cpu, policy->cpus);
return 0;
}
static int intel_pstate_cpu_exit(struct cpufreq_policy *policy)
{
return intel_pstate_exit_perf_limits(policy);
}
static struct cpufreq_driver intel_pstate_driver = {
.flags = CPUFREQ_CONST_LOOPS,
.verify = intel_pstate_verify_policy,
.setpolicy = intel_pstate_set_policy,
.get = intel_pstate_get,
.init = intel_pstate_cpu_init,
.exit = intel_pstate_cpu_exit,
.stop_cpu = intel_pstate_stop_cpu,
.name = "intel_pstate",
};
@ -1406,6 +1236,7 @@ static void copy_cpu_funcs(struct pstate_funcs *funcs)
}
#if IS_ENABLED(CONFIG_ACPI)
#include <acpi/processor.h>
static bool intel_pstate_no_acpi_pss(void)
{
@ -1601,9 +1432,6 @@ static int __init intel_pstate_setup(char *str)
force_load = 1;
if (!strcmp(str, "hwp_only"))
hwp_only = 1;
if (!strcmp(str, "no_acpi"))
no_acpi_perf = 1;
return 0;
}
early_param("intel_pstate", intel_pstate_setup);

2
drivers/crypto/qat/qat_common/adf_ctl_drv.c
View File

@ -198,7 +198,7 @@ static int adf_copy_key_value_data(struct adf_accel_dev *accel_dev,
goto out_err;
}
params_head = section_head->params;
params_head = section.params;
while (params_head) {
if (copy_from_user(&key_val, (void __user *)params_head,

20
drivers/dma/at_hdmac.c
View File

@ -729,8 +729,8 @@ atc_prep_dma_interleaved(struct dma_chan *chan,
return NULL;
dev_info(chan2dev(chan),
"%s: src=0x%08x, dest=0x%08x, numf=%d, frame_size=%d, flags=0x%lx\n",
__func__, xt->src_start, xt->dst_start, xt->numf,
"%s: src=%pad, dest=%pad, numf=%d, frame_size=%d, flags=0x%lx\n",
__func__, &xt->src_start, &xt->dst_start, xt->numf,
xt->frame_size, flags);
/*
@ -824,8 +824,8 @@ atc_prep_dma_memcpy(struct dma_chan *chan, dma_addr_t dest, dma_addr_t src,
u32 ctrla;
u32 ctrlb;
dev_vdbg(chan2dev(chan), "prep_dma_memcpy: d0x%x s0x%x l0x%zx f0x%lx\n",
dest, src, len, flags);
dev_vdbg(chan2dev(chan), "prep_dma_memcpy: d%pad s%pad l0x%zx f0x%lx\n",
&dest, &src, len, flags);
if (unlikely(!len)) {
dev_dbg(chan2dev(chan), "prep_dma_memcpy: length is zero!\n");
@ -938,8 +938,8 @@ atc_prep_dma_memset(struct dma_chan *chan, dma_addr_t dest, int value,
void __iomem *vaddr;
dma_addr_t paddr;
dev_vdbg(chan2dev(chan), "%s: d0x%x v0x%x l0x%zx f0x%lx\n", __func__,
dest, value, len, flags);
dev_vdbg(chan2dev(chan), "%s: d%pad v0x%x l0x%zx f0x%lx\n", __func__,
&dest, value, len, flags);
if (unlikely(!len)) {
dev_dbg(chan2dev(chan), "%s: length is zero!\n", __func__);
@ -1022,8 +1022,8 @@ atc_prep_dma_memset_sg(struct dma_chan *chan,
dma_addr_t dest = sg_dma_address(sg);
size_t len = sg_dma_len(sg);
dev_vdbg(chan2dev(chan), "%s: d0x%08x, l0x%zx\n",
__func__, dest, len);
dev_vdbg(chan2dev(chan), "%s: d%pad, l0x%zx\n",
__func__, &dest, len);
if (!is_dma_fill_aligned(chan->device, dest, 0, len)) {
dev_err(chan2dev(chan), "%s: buffer is not aligned\n",
@ -1439,9 +1439,9 @@ atc_prep_dma_cyclic(struct dma_chan *chan, dma_addr_t buf_addr, size_t buf_len,
unsigned int periods = buf_len / period_len;
unsigned int i;
dev_vdbg(chan2dev(chan), "prep_dma_cyclic: %s buf@0x%08x - %d (%d/%d)\n",
dev_vdbg(chan2dev(chan), "prep_dma_cyclic: %s buf@%pad - %d (%d/%d)\n",
direction == DMA_MEM_TO_DEV ? "TO DEVICE" : "FROM DEVICE",
buf_addr,
&buf_addr,
periods, buf_len, period_len);
if (unlikely(!atslave || !buf_len || !period_len)) {

6
drivers/dma/at_hdmac_regs.h
View File

@ -385,9 +385,9 @@ static void vdbg_dump_regs(struct at_dma_chan *atchan) {}
static void atc_dump_lli(struct at_dma_chan *atchan, struct at_lli *lli)
{
dev_crit(chan2dev(&atchan->chan_common),
" desc: s0x%x d0x%x ctrl0x%x:0x%x l0x%x\n",
lli->saddr, lli->daddr,
lli->ctrla, lli->ctrlb, lli->dscr);
" desc: s%pad d%pad ctrl0x%x:0x%x l0x%pad\n",
&lli->saddr, &lli->daddr,
lli->ctrla, lli->ctrlb, &lli->dscr);
}

20
drivers/dma/at_xdmac.c
View File

@ -920,8 +920,8 @@ at_xdmac_interleaved_queue_desc(struct dma_chan *chan,
desc->lld.mbr_cfg = chan_cc;
dev_dbg(chan2dev(chan),
"%s: lld: mbr_sa=0x%08x, mbr_da=0x%08x, mbr_ubc=0x%08x, mbr_cfg=0x%08x\n",
__func__, desc->lld.mbr_sa, desc->lld.mbr_da,
"%s: lld: mbr_sa=%pad, mbr_da=%pad, mbr_ubc=0x%08x, mbr_cfg=0x%08x\n",
__func__, &desc->lld.mbr_sa, &desc->lld.mbr_da,
desc->lld.mbr_ubc, desc->lld.mbr_cfg);
/* Chain lld. */
@ -953,8 +953,8 @@ at_xdmac_prep_interleaved(struct dma_chan *chan,
if ((xt->numf > 1) && (xt->frame_size > 1))
return NULL;
dev_dbg(chan2dev(chan), "%s: src=0x%08x, dest=0x%08x, numf=%d, frame_size=%d, flags=0x%lx\n",
__func__, xt->src_start, xt->dst_start, xt->numf,
dev_dbg(chan2dev(chan), "%s: src=%pad, dest=%pad, numf=%d, frame_size=%d, flags=0x%lx\n",
__func__, &xt->src_start, &xt->dst_start, xt->numf,
xt->frame_size, flags);
src_addr = xt->src_start;
@ -1179,8 +1179,8 @@ static struct at_xdmac_desc *at_xdmac_memset_create_desc(struct dma_chan *chan,
desc->lld.mbr_cfg = chan_cc;
dev_dbg(chan2dev(chan),
"%s: lld: mbr_da=0x%08x, mbr_ds=0x%08x, mbr_ubc=0x%08x, mbr_cfg=0x%08x\n",
__func__, desc->lld.mbr_da, desc->lld.mbr_ds, desc->lld.mbr_ubc,
"%s: lld: mbr_da=%pad, mbr_ds=%pad, mbr_ubc=0x%08x, mbr_cfg=0x%08x\n",
__func__, &desc->lld.mbr_da, &desc->lld.mbr_ds, desc->lld.mbr_ubc,
desc->lld.mbr_cfg);
return desc;
@ -1193,8 +1193,8 @@ at_xdmac_prep_dma_memset(struct dma_chan *chan, dma_addr_t dest, int value,
struct at_xdmac_chan *atchan = to_at_xdmac_chan(chan);
struct at_xdmac_desc *desc;
dev_dbg(chan2dev(chan), "%s: dest=0x%08x, len=%d, pattern=0x%x, flags=0x%lx\n",
__func__, dest, len, value, flags);
dev_dbg(chan2dev(chan), "%s: dest=%pad, len=%d, pattern=0x%x, flags=0x%lx\n",
__func__, &dest, len, value, flags);
if (unlikely(!len))
return NULL;
@ -1229,8 +1229,8 @@ at_xdmac_prep_dma_memset_sg(struct dma_chan *chan, struct scatterlist *sgl,
/* Prepare descriptors. */
for_each_sg(sgl, sg, sg_len, i) {
dev_dbg(chan2dev(chan), "%s: dest=0x%08x, len=%d, pattern=0x%x, flags=0x%lx\n",
__func__, sg_dma_address(sg), sg_dma_len(sg),
dev_dbg(chan2dev(chan), "%s: dest=%pad, len=%d, pattern=0x%x, flags=0x%lx\n",
__func__, &sg_dma_address(sg), sg_dma_len(sg),
value, flags);
desc = at_xdmac_memset_create_desc(chan, atchan,
sg_dma_address(sg),

4
drivers/dma/edma.c
View File

@ -107,7 +107,7 @@
/* CCCFG register */
#define GET_NUM_DMACH(x) (x & 0x7) /* bits 0-2 */
#define GET_NUM_QDMACH(x) (x & 0x70 >> 4) /* bits 4-6 */
#define GET_NUM_QDMACH(x) ((x & 0x70) >> 4) /* bits 4-6 */
#define GET_NUM_PAENTRY(x) ((x & 0x7000) >> 12) /* bits 12-14 */
#define GET_NUM_EVQUE(x) ((x & 0x70000) >> 16) /* bits 16-18 */
#define GET_NUM_REGN(x) ((x & 0x300000) >> 20) /* bits 20-21 */
@ -1565,7 +1565,7 @@ static void edma_tc_set_pm_state(struct edma_tc *tc, bool enable)
struct platform_device *tc_pdev;
int ret;
if (!tc)
if (!IS_ENABLED(CONFIG_OF) || !tc)
return;
tc_pdev = of_find_device_by_node(tc->node);

2
drivers/dma/imx-sdma.c
View File

@ -1462,7 +1462,7 @@ err_firmware:
#define EVENT_REMAP_CELLS 3
static int __init sdma_event_remap(struct sdma_engine *sdma)
static int sdma_event_remap(struct sdma_engine *sdma)
{
struct device_node *np = sdma->dev->of_node;
struct device_node *gpr_np = of_parse_phandle(np, "gpr", 0);

11
drivers/dma/sh/usb-dmac.c
View File

@ -679,8 +679,11 @@ static int usb_dmac_runtime_suspend(struct device *dev)
struct usb_dmac *dmac = dev_get_drvdata(dev);
int i;
for (i = 0; i < dmac->n_channels; ++i)
for (i = 0; i < dmac->n_channels; ++i) {
if (!dmac->channels[i].iomem)
break;
usb_dmac_chan_halt(&dmac->channels[i]);
}
return 0;
}
@ -799,11 +802,10 @@ static int usb_dmac_probe(struct platform_device *pdev)
ret = pm_runtime_get_sync(&pdev->dev);
if (ret < 0) {
dev_err(&pdev->dev, "runtime PM get sync failed (%d)\n", ret);
return ret;
goto error_pm;
}
ret = usb_dmac_init(dmac);
pm_runtime_put(&pdev->dev);
if (ret) {
dev_err(&pdev->dev, "failed to reset device\n");
@ -851,10 +853,13 @@ static int usb_dmac_probe(struct platform_device *pdev)
if (ret < 0)
goto error;
pm_runtime_put(&pdev->dev);
return 0;
error:
of_dma_controller_free(pdev->dev.of_node);
pm_runtime_put(&pdev->dev);
error_pm:
pm_runtime_disable(&pdev->dev);
return ret;
}

120
drivers/gpu/drm/amd/amdgpu/amdgpu.h
View File

@ -389,7 +389,6 @@ struct amdgpu_clock {
* Fences.
*/
struct amdgpu_fence_driver {
struct amdgpu_ring *ring;
uint64_t gpu_addr;
volatile uint32_t *cpu_addr;
/* sync_seq is protected by ring emission lock */
@ -398,7 +397,7 @@ struct amdgpu_fence_driver {
bool initialized;
struct amdgpu_irq_src *irq_src;
unsigned irq_type;
struct delayed_work lockup_work;
struct timer_list fallback_timer;
wait_queue_head_t fence_queue;
};
@ -917,8 +916,8 @@ struct amdgpu_ring {
#define AMDGPU_VM_FAULT_STOP_ALWAYS 2
struct amdgpu_vm_pt {
struct amdgpu_bo *bo;
uint64_t addr;
struct amdgpu_bo *bo;
uint64_t addr;
};
struct amdgpu_vm_id {
@ -926,8 +925,6 @@ struct amdgpu_vm_id {
uint64_t pd_gpu_addr;
/* last flushed PD/PT update */
struct fence *flushed_updates;
/* last use of vmid */
struct fence *last_id_use;
};
struct amdgpu_vm {
@ -957,24 +954,70 @@ struct amdgpu_vm {
/* for id and flush management per ring */
struct amdgpu_vm_id ids[AMDGPU_MAX_RINGS];
/* for interval tree */
spinlock_t it_lock;
};
struct amdgpu_vm_manager {
struct fence *active[AMDGPU_NUM_VM];
uint32_t max_pfn;
struct {
struct fence *active;
atomic_long_t owner;
} ids[AMDGPU_NUM_VM];
uint32_t max_pfn;
/* number of VMIDs */
unsigned nvm;
unsigned nvm;
/* vram base address for page table entry */
u64 vram_base_offset;
u64 vram_base_offset;
/* is vm enabled? */
bool enabled;
/* for hw to save the PD addr on suspend/resume */
uint32_t saved_table_addr[AMDGPU_NUM_VM];
bool enabled;
/* vm pte handling */
const struct amdgpu_vm_pte_funcs *vm_pte_funcs;
struct amdgpu_ring *vm_pte_funcs_ring;
};
void amdgpu_vm_manager_fini(struct amdgpu_device *adev);
int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm);
void amdgpu_vm_fini(struct amdgpu_device *adev, struct amdgpu_vm *vm);
struct amdgpu_bo_list_entry *amdgpu_vm_get_bos(struct amdgpu_device *adev,
struct amdgpu_vm *vm,
struct list_head *head);
int amdgpu_vm_grab_id(struct amdgpu_vm *vm, struct amdgpu_ring *ring,
struct amdgpu_sync *sync);
void amdgpu_vm_flush(struct amdgpu_ring *ring,
struct amdgpu_vm *vm,
struct fence *updates);
void amdgpu_vm_fence(struct amdgpu_device *adev,
struct amdgpu_vm *vm,
struct fence *fence);
uint64_t amdgpu_vm_map_gart(struct amdgpu_device *adev, uint64_t addr);
int amdgpu_vm_update_page_directory(struct amdgpu_device *adev,
struct amdgpu_vm *vm);
int amdgpu_vm_clear_freed(struct amdgpu_device *adev,
struct amdgpu_vm *vm);
int amdgpu_vm_clear_invalids(struct amdgpu_device *adev, struct amdgpu_vm *vm,
struct amdgpu_sync *sync);
int amdgpu_vm_bo_update(struct amdgpu_device *adev,
struct amdgpu_bo_va *bo_va,
struct ttm_mem_reg *mem);
void amdgpu_vm_bo_invalidate(struct amdgpu_device *adev,
struct amdgpu_bo *bo);
struct amdgpu_bo_va *amdgpu_vm_bo_find(struct amdgpu_vm *vm,
struct amdgpu_bo *bo);
struct amdgpu_bo_va *amdgpu_vm_bo_add(struct amdgpu_device *adev,
struct amdgpu_vm *vm,
struct amdgpu_bo *bo);
int amdgpu_vm_bo_map(struct amdgpu_device *adev,
struct amdgpu_bo_va *bo_va,
uint64_t addr, uint64_t offset,
uint64_t size, uint32_t flags);
int amdgpu_vm_bo_unmap(struct amdgpu_device *adev,
struct amdgpu_bo_va *bo_va,
uint64_t addr);
void amdgpu_vm_bo_rmv(struct amdgpu_device *adev,
struct amdgpu_bo_va *bo_va);
int amdgpu_vm_free_job(struct amdgpu_job *job);
/*
* context related structures
*/
@ -1211,6 +1254,7 @@ struct amdgpu_cs_parser {
/* relocations */
struct amdgpu_bo_list_entry *vm_bos;
struct list_head validated;
struct fence *fence;
struct amdgpu_ib *ibs;
uint32_t num_ibs;
@ -1226,7 +1270,7 @@ struct amdgpu_job {
struct amdgpu_device *adev;
struct amdgpu_ib *ibs;
uint32_t num_ibs;
struct mutex job_lock;
void *owner;
struct amdgpu_user_fence uf;
int (*free_job)(struct amdgpu_job *job);
};
@ -2257,11 +2301,6 @@ void amdgpu_pci_config_reset(struct amdgpu_device *adev);
bool amdgpu_card_posted(struct amdgpu_device *adev);
void amdgpu_update_display_priority(struct amdgpu_device *adev);
bool amdgpu_boot_test_post_card(struct amdgpu_device *adev);
struct amdgpu_cs_parser *amdgpu_cs_parser_create(struct amdgpu_device *adev,
struct drm_file *filp,
struct amdgpu_ctx *ctx,
struct amdgpu_ib *ibs,
uint32_t num_ibs);
int amdgpu_cs_parser_init(struct amdgpu_cs_parser *p, void *data);
int amdgpu_cs_get_ring(struct amdgpu_device *adev, u32 ip_type,
@ -2318,49 +2357,6 @@ int amdgpu_get_vblank_timestamp_kms(struct drm_device *dev, unsigned int pipe,
long amdgpu_kms_compat_ioctl(struct file *filp, unsigned int cmd,
unsigned long arg);
/*
* vm
*/
int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm);
void amdgpu_vm_fini(struct amdgpu_device *adev, struct amdgpu_vm *vm);
struct amdgpu_bo_list_entry *amdgpu_vm_get_bos(struct amdgpu_device *adev,
struct amdgpu_vm *vm,
struct list_head *head);
int amdgpu_vm_grab_id(struct amdgpu_vm *vm, struct amdgpu_ring *ring,
struct amdgpu_sync *sync);
void amdgpu_vm_flush(struct amdgpu_ring *ring,
struct amdgpu_vm *vm,
struct fence *updates);
void amdgpu_vm_fence(struct amdgpu_device *adev,
struct amdgpu_vm *vm,
struct amdgpu_fence *fence);
uint64_t amdgpu_vm_map_gart(struct amdgpu_device *adev, uint64_t addr);
int amdgpu_vm_update_page_directory(struct amdgpu_device *adev,
struct amdgpu_vm *vm);
int amdgpu_vm_clear_freed(struct amdgpu_device *adev,
struct amdgpu_vm *vm);
int amdgpu_vm_clear_invalids(struct amdgpu_device *adev,
struct amdgpu_vm *vm, struct amdgpu_sync *sync);
int amdgpu_vm_bo_update(struct amdgpu_device *adev,
struct amdgpu_bo_va *bo_va,
struct ttm_mem_reg *mem);
void amdgpu_vm_bo_invalidate(struct amdgpu_device *adev,
struct amdgpu_bo *bo);
struct amdgpu_bo_va *amdgpu_vm_bo_find(struct amdgpu_vm *vm,
struct amdgpu_bo *bo);
struct amdgpu_bo_va *amdgpu_vm_bo_add(struct amdgpu_device *adev,
struct amdgpu_vm *vm,
struct amdgpu_bo *bo);
int amdgpu_vm_bo_map(struct amdgpu_device *adev,
struct amdgpu_bo_va *bo_va,
uint64_t addr, uint64_t offset,
uint64_t size, uint32_t flags);
int amdgpu_vm_bo_unmap(struct amdgpu_device *adev,
struct amdgpu_bo_va *bo_va,
uint64_t addr);
void amdgpu_vm_bo_rmv(struct amdgpu_device *adev,
struct amdgpu_bo_va *bo_va);
int amdgpu_vm_free_job(struct amdgpu_job *job);
/*
* functions used by amdgpu_encoder.c
*/

177
drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
View File

@ -127,30 +127,6 @@ int amdgpu_cs_get_ring(struct amdgpu_device *adev, u32 ip_type,
return 0;
}
struct amdgpu_cs_parser *amdgpu_cs_parser_create(struct amdgpu_device *adev,
struct drm_file *filp,
struct amdgpu_ctx *ctx,
struct amdgpu_ib *ibs,
uint32_t num_ibs)
{
struct amdgpu_cs_parser *parser;
int i;
parser = kzalloc(sizeof(struct amdgpu_cs_parser), GFP_KERNEL);
if (!parser)
return NULL;
parser->adev = adev;
parser->filp = filp;
parser->ctx = ctx;
parser->ibs = ibs;
parser->num_ibs = num_ibs;
for (i = 0; i < num_ibs; i++)
ibs[i].ctx = ctx;
return parser;
}
int amdgpu_cs_parser_init(struct amdgpu_cs_parser *p, void *data)
{
union drm_amdgpu_cs *cs = data;
@ -463,8 +439,18 @@ static int cmp_size_smaller_first(void *priv, struct list_head *a,
return (int)la->robj->tbo.num_pages - (int)lb->robj->tbo.num_pages;
}
static void amdgpu_cs_parser_fini_early(struct amdgpu_cs_parser *parser, int error, bool backoff)
/**
* cs_parser_fini() - clean parser states
* @parser: parser structure holding parsing context.
* @error: error number
*
* If error is set than unvalidate buffer, otherwise just free memory
* used by parsing context.
**/
static void amdgpu_cs_parser_fini(struct amdgpu_cs_parser *parser, int error, bool backoff)
{
unsigned i;
if (!error) {
/* Sort the buffer list from the smallest to largest buffer,
* which affects the order of buffers in the LRU list.
@ -479,17 +465,14 @@ static void amdgpu_cs_parser_fini_early(struct amdgpu_cs_parser *parser, int err
list_sort(NULL, &parser->validated, cmp_size_smaller_first);
ttm_eu_fence_buffer_objects(&parser->ticket,
&parser->validated,
&parser->ibs[parser->num_ibs-1].fence->base);
&parser->validated,
parser->fence);
} else if (backoff) {
ttm_eu_backoff_reservation(&parser->ticket,
&parser->validated);
}
}
fence_put(parser->fence);
static void amdgpu_cs_parser_fini_late(struct amdgpu_cs_parser *parser)
{
unsigned i;
if (parser->ctx)
amdgpu_ctx_put(parser->ctx);
if (parser->bo_list)
@ -499,31 +482,12 @@ static void amdgpu_cs_parser_fini_late(struct amdgpu_cs_parser *parser)
for (i = 0; i < parser->nchunks; i++)
drm_free_large(parser->chunks[i].kdata);
kfree(parser->chunks);
if (!amdgpu_enable_scheduler)
{
if (parser->ibs)
for (i = 0; i < parser->num_ibs; i++)
amdgpu_ib_free(parser->adev, &parser->ibs[i]);
kfree(parser->ibs);
if (parser->uf.bo)
drm_gem_object_unreference_unlocked(&parser->uf.bo->gem_base);
}
kfree(parser);
}
/**
* cs_parser_fini() - clean parser states
* @parser: parser structure holding parsing context.
* @error: error number
*
* If error is set than unvalidate buffer, otherwise just free memory
* used by parsing context.
**/
static void amdgpu_cs_parser_fini(struct amdgpu_cs_parser *parser, int error, bool backoff)
{
amdgpu_cs_parser_fini_early(parser, error, backoff);
amdgpu_cs_parser_fini_late(parser);
if (parser->ibs)
for (i = 0; i < parser->num_ibs; i++)
amdgpu_ib_free(parser->adev, &parser->ibs[i]);
kfree(parser->ibs);
if (parser->uf.bo)
drm_gem_object_unreference_unlocked(&parser->uf.bo->gem_base);
}
static int amdgpu_bo_vm_update_pte(struct amdgpu_cs_parser *p,
@ -610,15 +574,9 @@ static int amdgpu_cs_ib_vm_chunk(struct amdgpu_device *adev,
}
r = amdgpu_bo_vm_update_pte(parser, vm);
if (r) {
goto out;
}
amdgpu_cs_sync_rings(parser);
if (!amdgpu_enable_scheduler)
r = amdgpu_ib_schedule(adev, parser->num_ibs, parser->ibs,
parser->filp);
if (!r)
amdgpu_cs_sync_rings(parser);
out:
return r;
}
@ -828,36 +786,36 @@ int amdgpu_cs_ioctl(struct drm_device *dev, void *data, struct drm_file *filp)
union drm_amdgpu_cs *cs = data;
struct amdgpu_fpriv *fpriv = filp->driver_priv;
struct amdgpu_vm *vm = &fpriv->vm;
struct amdgpu_cs_parser *parser;
struct amdgpu_cs_parser parser = {};
bool reserved_buffers = false;
int i, r;
if (!adev->accel_working)
return -EBUSY;
parser = amdgpu_cs_parser_create(adev, filp, NULL, NULL, 0);
if (!parser)
return -ENOMEM;
r = amdgpu_cs_parser_init(parser, data);
parser.adev = adev;
parser.filp = filp;
r = amdgpu_cs_parser_init(&parser, data);
if (r) {
DRM_ERROR("Failed to initialize parser !\n");
amdgpu_cs_parser_fini(parser, r, false);
amdgpu_cs_parser_fini(&parser, r, false);
r = amdgpu_cs_handle_lockup(adev, r);
return r;
}
mutex_lock(&vm->mutex);
r = amdgpu_cs_parser_relocs(parser);
r = amdgpu_cs_parser_relocs(&parser);
if (r == -ENOMEM)
DRM_ERROR("Not enough memory for command submission!\n");
else if (r && r != -ERESTARTSYS)
DRM_ERROR("Failed to process the buffer list %d!\n", r);
else if (!r) {
reserved_buffers = true;
r = amdgpu_cs_ib_fill(adev, parser);
r = amdgpu_cs_ib_fill(adev, &parser);
}
if (!r) {
r = amdgpu_cs_dependencies(adev, parser);
r = amdgpu_cs_dependencies(adev, &parser);
if (r)
DRM_ERROR("Failed in the dependencies handling %d!\n", r);
}
@ -865,62 +823,71 @@ int amdgpu_cs_ioctl(struct drm_device *dev, void *data, struct drm_file *filp)
if (r)
goto out;
for (i = 0; i < parser->num_ibs; i++)
trace_amdgpu_cs(parser, i);
for (i = 0; i < parser.num_ibs; i++)
trace_amdgpu_cs(&parser, i);
r = amdgpu_cs_ib_vm_chunk(adev, parser);
r = amdgpu_cs_ib_vm_chunk(adev, &parser);
if (r)
goto out;
if (amdgpu_enable_scheduler && parser->num_ibs) {
if (amdgpu_enable_scheduler && parser.num_ibs) {
struct amdgpu_ring * ring = parser.ibs->ring;
struct amd_sched_fence *fence;
struct amdgpu_job *job;
struct amdgpu_ring * ring = parser->ibs->ring;
job = kzalloc(sizeof(struct amdgpu_job), GFP_KERNEL);
if (!job) {
r = -ENOMEM;
goto out;
}
job->base.sched = &ring->sched;
job->base.s_entity = &parser->ctx->rings[ring->idx].entity;
job->adev = parser->adev;
job->ibs = parser->ibs;
job->num_ibs = parser->num_ibs;
job->base.owner = parser->filp;
mutex_init(&job->job_lock);
job->base.s_entity = &parser.ctx->rings[ring->idx].entity;
job->adev = parser.adev;
job->owner = parser.filp;
job->free_job = amdgpu_cs_free_job;
job->ibs = parser.ibs;
job->num_ibs = parser.num_ibs;
parser.ibs = NULL;
parser.num_ibs = 0;
if (job->ibs[job->num_ibs - 1].user) {
memcpy(&job->uf, &parser->uf,
sizeof(struct amdgpu_user_fence));
job->uf = parser.uf;
job->ibs[job->num_ibs - 1].user = &job->uf;
parser.uf.bo = NULL;
}
job->free_job = amdgpu_cs_free_job;
mutex_lock(&job->job_lock);
r = amd_sched_entity_push_job(&job->base);
if (r) {
mutex_unlock(&job->job_lock);
fence = amd_sched_fence_create(job->base.s_entity,
parser.filp);
if (!fence) {
r = -ENOMEM;
amdgpu_cs_free_job(job);
kfree(job);
goto out;
}
cs->out.handle =
amdgpu_ctx_add_fence(parser->ctx, ring,
&job->base.s_fence->base);
parser->ibs[parser->num_ibs - 1].sequence = cs->out.handle;
job->base.s_fence = fence;
parser.fence = fence_get(&fence->base);
list_sort(NULL, &parser->validated, cmp_size_smaller_first);
ttm_eu_fence_buffer_objects(&parser->ticket,
&parser->validated,
&job->base.s_fence->base);
cs->out.handle = amdgpu_ctx_add_fence(parser.ctx, ring,
&fence->base);
job->ibs[job->num_ibs - 1].sequence = cs->out.handle;
mutex_unlock(&job->job_lock);
amdgpu_cs_parser_fini_late(parser);
mutex_unlock(&vm->mutex);
return 0;
trace_amdgpu_cs_ioctl(job);
amd_sched_entity_push_job(&job->base);
} else {
struct amdgpu_fence *fence;
r = amdgpu_ib_schedule(adev, parser.num_ibs, parser.ibs,
parser.filp);
fence = parser.ibs[parser.num_ibs - 1].fence;
parser.fence = fence_get(&fence->base);
cs->out.handle = parser.ibs[parser.num_ibs - 1].sequence;
}
cs->out.handle = parser->ibs[parser->num_ibs - 1].sequence;
out:
amdgpu_cs_parser_fini(parser, r, reserved_buffers);
amdgpu_cs_parser_fini(&parser, r, reserved_buffers);
mutex_unlock(&vm->mutex);
r = amdgpu_cs_handle_lockup(adev, r);
return r;

103
drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c
View File

@ -47,6 +47,9 @@
* that the the relevant GPU caches have been flushed.
*/
static struct kmem_cache *amdgpu_fence_slab;
static atomic_t amdgpu_fence_slab_ref = ATOMIC_INIT(0);
/**
* amdgpu_fence_write - write a fence value
*
@ -84,24 +87,6 @@ static u32 amdgpu_fence_read(struct amdgpu_ring *ring)
return seq;
}
/**
* amdgpu_fence_schedule_check - schedule lockup check
*
* @ring: pointer to struct amdgpu_ring
*
* Queues a delayed work item to check for lockups.
*/
static void amdgpu_fence_schedule_check(struct amdgpu_ring *ring)
{
/*
* Do not reset the timer here with mod_delayed_work,
* this can livelock in an interaction with TTM delayed destroy.
*/
queue_delayed_work(system_power_efficient_wq,
&ring->fence_drv.lockup_work,
AMDGPU_FENCE_JIFFIES_TIMEOUT);
}
/**
* amdgpu_fence_emit - emit a fence on the requested ring
*
@ -118,7 +103,7 @@ int amdgpu_fence_emit(struct amdgpu_ring *ring, void *owner,
struct amdgpu_device *adev = ring->adev;
/* we are protected by the ring emission mutex */
*fence = kmalloc(sizeof(struct amdgpu_fence), GFP_KERNEL);
*fence = kmem_cache_alloc(amdgpu_fence_slab, GFP_KERNEL);
if ((*fence) == NULL) {
return -ENOMEM;
}
@ -132,10 +117,22 @@ int amdgpu_fence_emit(struct amdgpu_ring *ring, void *owner,
amdgpu_ring_emit_fence(ring, ring->fence_drv.gpu_addr,
(*fence)->seq,
AMDGPU_FENCE_FLAG_INT);
trace_amdgpu_fence_emit(ring->adev->ddev, ring->idx, (*fence)->seq);
return 0;
}
/**
* amdgpu_fence_schedule_fallback - schedule fallback check
*
* @ring: pointer to struct amdgpu_ring
*
* Start a timer as fallback to our interrupts.
*/
static void amdgpu_fence_schedule_fallback(struct amdgpu_ring *ring)
{
mod_timer(&ring->fence_drv.fallback_timer,
jiffies + AMDGPU_FENCE_JIFFIES_TIMEOUT);
}
/**
* amdgpu_fence_activity - check for fence activity
*
@ -202,32 +199,11 @@ static bool amdgpu_fence_activity(struct amdgpu_ring *ring)
} while (atomic64_xchg(&ring->fence_drv.last_seq, seq) > seq);
if (seq < last_emitted)
amdgpu_fence_schedule_check(ring);
amdgpu_fence_schedule_fallback(ring);
return wake;
}
/**
* amdgpu_fence_check_lockup - check for hardware lockup
*
* @work: delayed work item
*
* Checks for fence activity and if there is none probe
* the hardware if a lockup occured.
*/
static void amdgpu_fence_check_lockup(struct work_struct *work)
{
struct amdgpu_fence_driver *fence_drv;
struct amdgpu_ring *ring;
fence_drv = container_of(work, struct amdgpu_fence_driver,
lockup_work.work);
ring = fence_drv->ring;
if (amdgpu_fence_activity(ring))
wake_up_all(&ring->fence_drv.fence_queue);
}
/**
* amdgpu_fence_process - process a fence
*
@ -243,6 +219,20 @@ void amdgpu_fence_process(struct amdgpu_ring *ring)
wake_up_all(&ring->fence_drv.fence_queue);
}
/**
* amdgpu_fence_fallback - fallback for hardware interrupts
*
* @work: delayed work item
*
* Checks for fence activity.
*/
static void amdgpu_fence_fallback(unsigned long arg)
{
struct amdgpu_ring *ring = (void *)arg;
amdgpu_fence_process(ring);
}
/**
* amdgpu_fence_seq_signaled - check if a fence sequence number has signaled
*
@ -290,7 +280,7 @@ static int amdgpu_fence_ring_wait_seq(struct amdgpu_ring *ring, uint64_t seq)
if (atomic64_read(&ring->fence_drv.last_seq) >= seq)
return 0;
amdgpu_fence_schedule_check(ring);
amdgpu_fence_schedule_fallback(ring);
wait_event(ring->fence_drv.fence_queue, (
(signaled = amdgpu_fence_seq_signaled(ring, seq))));
@ -491,9 +481,8 @@ int amdgpu_fence_driver_init_ring(struct amdgpu_ring *ring)
atomic64_set(&ring->fence_drv.last_seq, 0);
ring->fence_drv.initialized = false;
INIT_DELAYED_WORK(&ring->fence_drv.lockup_work,
amdgpu_fence_check_lockup);
ring->fence_drv.ring = ring;
setup_timer(&ring->fence_drv.fallback_timer, amdgpu_fence_fallback,
(unsigned long)ring);
init_waitqueue_head(&ring->fence_drv.fence_queue);
@ -536,6 +525,13 @@ int amdgpu_fence_driver_init_ring(struct amdgpu_ring *ring)
*/
int amdgpu_fence_driver_init(struct amdgpu_device *adev)
{
if (atomic_inc_return(&amdgpu_fence_slab_ref) == 1) {
amdgpu_fence_slab = kmem_cache_create(
"amdgpu_fence", sizeof(struct amdgpu_fence), 0,
SLAB_HWCACHE_ALIGN, NULL);
if (!amdgpu_fence_slab)
return -ENOMEM;
}
if (amdgpu_debugfs_fence_init(adev))
dev_err(adev->dev, "fence debugfs file creation failed\n");
@ -554,9 +550,12 @@ void amdgpu_fence_driver_fini(struct amdgpu_device *adev)
{
int i, r;
if (atomic_dec_and_test(&amdgpu_fence_slab_ref))
kmem_cache_destroy(amdgpu_fence_slab);
mutex_lock(&adev->ring_lock);
for (i = 0; i < AMDGPU_MAX_RINGS; i++) {
struct amdgpu_ring *ring = adev->rings[i];
if (!ring || !ring->fence_drv.initialized)
continue;
r = amdgpu_fence_wait_empty(ring);
@ -568,6 +567,7 @@ void amdgpu_fence_driver_fini(struct amdgpu_device *adev)
amdgpu_irq_put(adev, ring->fence_drv.irq_src,
ring->fence_drv.irq_type);
amd_sched_fini(&ring->sched);
del_timer_sync(&ring->fence_drv.fallback_timer);
ring->fence_drv.initialized = false;
}
mutex_unlock(&adev->ring_lock);
@ -751,18 +751,25 @@ static bool amdgpu_fence_enable_signaling(struct fence *f)
fence->fence_wake.func = amdgpu_fence_check_signaled;
__add_wait_queue(&ring->fence_drv.fence_queue, &fence->fence_wake);
fence_get(f);
amdgpu_fence_schedule_check(ring);
if (!timer_pending(&ring->fence_drv.fallback_timer))
amdgpu_fence_schedule_fallback(ring);
FENCE_TRACE(&fence->base, "armed on ring %i!\n", ring->idx);
return true;
}
static void amdgpu_fence_release(struct fence *f)
{
struct amdgpu_fence *fence = to_amdgpu_fence(f);
kmem_cache_free(amdgpu_fence_slab, fence);
}
const struct fence_ops amdgpu_fence_ops = {
.get_driver_name = amdgpu_fence_get_driver_name,
.get_timeline_name = amdgpu_fence_get_timeline_name,
.enable_signaling = amdgpu_fence_enable_signaling,
.signaled = amdgpu_fence_is_signaled,
.wait = fence_default_wait,
.release = NULL,
.release = amdgpu_fence_release,
};
/*

24
drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
View File

@ -483,6 +483,9 @@ static void amdgpu_gem_va_update_vm(struct amdgpu_device *adev,
if (domain == AMDGPU_GEM_DOMAIN_CPU)
goto error_unreserve;
}
r = amdgpu_vm_update_page_directory(adev, bo_va->vm);
if (r)
goto error_unreserve;
r = amdgpu_vm_clear_freed(adev, bo_va->vm);
if (r)
@ -512,6 +515,9 @@ int amdgpu_gem_va_ioctl(struct drm_device *dev, void *data,
struct amdgpu_fpriv *fpriv = filp->driver_priv;
struct amdgpu_bo *rbo;
struct amdgpu_bo_va *bo_va;
struct ttm_validate_buffer tv, tv_pd;
struct ww_acquire_ctx ticket;
struct list_head list, duplicates;
uint32_t invalid_flags, va_flags = 0;
int r = 0;
@ -549,7 +555,18 @@ int amdgpu_gem_va_ioctl(struct drm_device *dev, void *data,
return -ENOENT;
mutex_lock(&fpriv->vm.mutex);
rbo = gem_to_amdgpu_bo(gobj);
r = amdgpu_bo_reserve(rbo, false);
INIT_LIST_HEAD(&list);
INIT_LIST_HEAD(&duplicates);
tv.bo = &rbo->tbo;
tv.shared = true;
list_add(&tv.head, &list);
if (args->operation == AMDGPU_VA_OP_MAP) {
tv_pd.bo = &fpriv->vm.page_directory->tbo;
tv_pd.shared = true;
list_add(&tv_pd.head, &list);
}
r = ttm_eu_reserve_buffers(&ticket, &list, true, &duplicates);
if (r) {
mutex_unlock(&fpriv->vm.mutex);
drm_gem_object_unreference_unlocked(gobj);
@ -558,7 +575,8 @@ int amdgpu_gem_va_ioctl(struct drm_device *dev, void *data,
bo_va = amdgpu_vm_bo_find(&fpriv->vm, rbo);
if (!bo_va) {
amdgpu_bo_unreserve(rbo);
ttm_eu_backoff_reservation(&ticket, &list);
drm_gem_object_unreference_unlocked(gobj);
mutex_unlock(&fpriv->vm.mutex);
return -ENOENT;
}
@ -581,7 +599,7 @@ int amdgpu_gem_va_ioctl(struct drm_device *dev, void *data,
default:
break;
}
ttm_eu_backoff_reservation(&ticket, &list);
if (!r && !(args->flags & AMDGPU_VM_DELAY_UPDATE))
amdgpu_gem_va_update_vm(adev, bo_va, args->operation);
mutex_unlock(&fpriv->vm.mutex);

4
drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c
View File

@ -62,7 +62,7 @@ int amdgpu_ib_get(struct amdgpu_ring *ring, struct amdgpu_vm *vm,
int r;
if (size) {
r = amdgpu_sa_bo_new(adev, &adev->ring_tmp_bo,
r = amdgpu_sa_bo_new(&adev->ring_tmp_bo,
&ib->sa_bo, size, 256);
if (r) {
dev_err(adev->dev, "failed to get a new IB (%d)\n", r);
@ -216,7 +216,7 @@ int amdgpu_ib_schedule(struct amdgpu_device *adev, unsigned num_ibs,
}
if (ib->vm)
amdgpu_vm_fence(adev, ib->vm, ib->fence);
amdgpu_vm_fence(adev, ib->vm, &ib->fence->base);
amdgpu_ring_unlock_commit(ring);
return 0;

7
drivers/gpu/drm/amd/amdgpu/amdgpu_object.h
View File

@ -189,10 +189,9 @@ int amdgpu_sa_bo_manager_start(struct amdgpu_device *adev,
struct amdgpu_sa_manager *sa_manager);
int amdgpu_sa_bo_manager_suspend(struct amdgpu_device *adev,
struct amdgpu_sa_manager *sa_manager);
int amdgpu_sa_bo_new(struct amdgpu_device *adev,
struct amdgpu_sa_manager *sa_manager,
struct amdgpu_sa_bo **sa_bo,
unsigned size, unsigned align);
int amdgpu_sa_bo_new(struct amdgpu_sa_manager *sa_manager,
struct amdgpu_sa_bo **sa_bo,
unsigned size, unsigned align);
void amdgpu_sa_bo_free(struct amdgpu_device *adev,
struct amdgpu_sa_bo **sa_bo,
struct fence *fence);

3
drivers/gpu/drm/amd/amdgpu/amdgpu_sa.c
View File

@ -311,8 +311,7 @@ static bool amdgpu_sa_bo_next_hole(struct amdgpu_sa_manager *sa_manager,
return false;
}
int amdgpu_sa_bo_new(struct amdgpu_device *adev,
struct amdgpu_sa_manager *sa_manager,
int amdgpu_sa_bo_new(struct amdgpu_sa_manager *sa_manager,
struct amdgpu_sa_bo **sa_bo,
unsigned size, unsigned align)
{

30
drivers/gpu/drm/amd/amdgpu/amdgpu_sched.c
View File

@ -26,6 +26,7 @@
#include <linux/sched.h>
#include <drm/drmP.h>
#include "amdgpu.h"
#include "amdgpu_trace.h"
static struct fence *amdgpu_sched_dependency(struct amd_sched_job *sched_job)
{
@ -44,11 +45,8 @@ static struct fence *amdgpu_sched_run_job(struct amd_sched_job *sched_job)
return NULL;
}
job = to_amdgpu_job(sched_job);
mutex_lock(&job->job_lock);
r = amdgpu_ib_schedule(job->adev,
job->num_ibs,
job->ibs,
job->base.owner);
trace_amdgpu_sched_run_job(job);
r = amdgpu_ib_schedule(job->adev, job->num_ibs, job->ibs, job->owner);
if (r) {
DRM_ERROR("Error scheduling IBs (%d)\n", r);
goto err;
@ -61,8 +59,6 @@ err:
if (job->free_job)
job->free_job(job);
mutex_unlock(&job->job_lock);
fence_put(&job->base.s_fence->base);
kfree(job);
return fence ? &fence->base : NULL;
}
@ -88,21 +84,19 @@ int amdgpu_sched_ib_submit_kernel_helper(struct amdgpu_device *adev,
return -ENOMEM;
job->base.sched = &ring->sched;
job->base.s_entity = &adev->kernel_ctx.rings[ring->idx].entity;
job->base.s_fence = amd_sched_fence_create(job->base.s_entity, owner);
if (!job->base.s_fence) {
kfree(job);
return -ENOMEM;
}
*f = fence_get(&job->base.s_fence->base);
job->adev = adev;
job->ibs = ibs;
job->num_ibs = num_ibs;
job->base.owner = owner;
mutex_init(&job->job_lock);
job->owner = owner;
job->free_job = free_job;
mutex_lock(&job->job_lock);
r = amd_sched_entity_push_job(&job->base);
if (r) {
mutex_unlock(&job->job_lock);
kfree(job);
return r;
}
*f = fence_get(&job->base.s_fence->base);
mutex_unlock(&job->job_lock);
amd_sched_entity_push_job(&job->base);
} else {
r = amdgpu_ib_schedule(adev, num_ibs, ibs, owner);
if (r)

2
drivers/gpu/drm/amd/amdgpu/amdgpu_semaphore.c
View File

@ -40,7 +40,7 @@ int amdgpu_semaphore_create(struct amdgpu_device *adev,
if (*semaphore == NULL) {
return -ENOMEM;
}
r = amdgpu_sa_bo_new(adev, &adev->ring_tmp_bo,
r = amdgpu_sa_bo_new(&adev->ring_tmp_bo,
&(*semaphore)->sa_bo, 8, 8);
if (r) {
kfree(*semaphore);

10
drivers/gpu/drm/amd/amdgpu/amdgpu_sync.c
View File

@ -302,8 +302,14 @@ int amdgpu_sync_rings(struct amdgpu_sync *sync,
return -EINVAL;
}
if (amdgpu_enable_scheduler || !amdgpu_enable_semaphores ||
(count >= AMDGPU_NUM_SYNCS)) {
if (amdgpu_enable_scheduler || !amdgpu_enable_semaphores) {
r = fence_wait(&fence->base, true);
if (r)
return r;
continue;
}
if (count >= AMDGPU_NUM_SYNCS) {
/* not enough room, wait manually */
r = fence_wait(&fence->base, false);
if (r)

94
drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h
View File

@ -48,6 +48,57 @@ TRACE_EVENT(amdgpu_cs,
__entry->fences)
);
TRACE_EVENT(amdgpu_cs_ioctl,
TP_PROTO(struct amdgpu_job *job),
TP_ARGS(job),
TP_STRUCT__entry(
__field(struct amdgpu_device *, adev)
__field(struct amd_sched_job *, sched_job)
__field(struct amdgpu_ib *, ib)
__field(struct fence *, fence)
__field(char *, ring_name)
__field(u32, num_ibs)
),
TP_fast_assign(
__entry->adev = job->adev;
__entry->sched_job = &job->base;
__entry->ib = job->ibs;
__entry->fence = &job->base.s_fence->base;
__entry->ring_name = job->ibs[0].ring->name;
__entry->num_ibs = job->num_ibs;
),
TP_printk("adev=%p, sched_job=%p, first ib=%p, sched fence=%p, ring name:%s, num_ibs:%u",
__entry->adev, __entry->sched_job, __entry->ib,
__entry->fence, __entry->ring_name, __entry->num_ibs)
);
TRACE_EVENT(amdgpu_sched_run_job,
TP_PROTO(struct amdgpu_job *job),
TP_ARGS(job),
TP_STRUCT__entry(
__field(struct amdgpu_device *, adev)
__field(struct amd_sched_job *, sched_job)
__field(struct amdgpu_ib *, ib)
__field(struct fence *, fence)
__field(char *, ring_name)
__field(u32, num_ibs)
),
TP_fast_assign(
__entry->adev = job->adev;
__entry->sched_job = &job->base;
__entry->ib = job->ibs;
__entry->fence = &job->base.s_fence->base;
__entry->ring_name = job->ibs[0].ring->name;
__entry->num_ibs = job->num_ibs;
),
TP_printk("adev=%p, sched_job=%p, first ib=%p, sched fence=%p, ring name:%s, num_ibs:%u",
__entry->adev, __entry->sched_job, __entry->ib,
__entry->fence, __entry->ring_name, __entry->num_ibs)
);
TRACE_EVENT(amdgpu_vm_grab_id,
TP_PROTO(unsigned vmid, int ring),
TP_ARGS(vmid, ring),
@ -196,49 +247,6 @@ TRACE_EVENT(amdgpu_bo_list_set,
TP_printk("list=%p, bo=%p", __entry->list, __entry->bo)
);
DECLARE_EVENT_CLASS(amdgpu_fence_request,
TP_PROTO(struct drm_device *dev, int ring, u32 seqno),
TP_ARGS(dev, ring, seqno),
TP_STRUCT__entry(
__field(u32, dev)
__field(int, ring)
__field(u32, seqno)
),
TP_fast_assign(
__entry->dev = dev->primary->index;
__entry->ring = ring;
__entry->seqno = seqno;
),
TP_printk("dev=%u, ring=%d, seqno=%u",
__entry->dev, __entry->ring, __entry->seqno)
);
DEFINE_EVENT(amdgpu_fence_request, amdgpu_fence_emit,
TP_PROTO(struct drm_device *dev, int ring, u32 seqno),
TP_ARGS(dev, ring, seqno)
);
DEFINE_EVENT(amdgpu_fence_request, amdgpu_fence_wait_begin,
TP_PROTO(struct drm_device *dev, int ring, u32 seqno),
TP_ARGS(dev, ring, seqno)
);
DEFINE_EVENT(amdgpu_fence_request, amdgpu_fence_wait_end,
TP_PROTO(struct drm_device *dev, int ring, u32 seqno),
TP_ARGS(dev, ring, seqno)
);
DECLARE_EVENT_CLASS(amdgpu_semaphore_request,
TP_PROTO(int ring, struct amdgpu_semaphore *sem),

6
drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
View File

@ -1073,10 +1073,10 @@ static int amdgpu_mm_dump_table(struct seq_file *m, void *data)
ret = drm_mm_dump_table(m, mm);
spin_unlock(&glob->lru_lock);
if (ttm_pl == TTM_PL_VRAM)
seq_printf(m, "man size:%llu pages, ram usage:%luMB, vis usage:%luMB\n",
seq_printf(m, "man size:%llu pages, ram usage:%lluMB, vis usage:%lluMB\n",
adev->mman.bdev.man[ttm_pl].size,
atomic64_read(&adev->vram_usage) >> 20,
atomic64_read(&adev->vram_vis_usage) >> 20);
(u64)atomic64_read(&adev->vram_usage) >> 20,
(u64)atomic64_read(&adev->vram_vis_usage) >> 20);
return ret;
}

138
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
View File

@ -143,10 +143,15 @@ int amdgpu_vm_grab_id(struct amdgpu_vm *vm, struct amdgpu_ring *ring,
unsigned i;
/* check if the id is still valid */
if (vm_id->id && vm_id->last_id_use &&
vm_id->last_id_use == adev->vm_manager.active[vm_id->id]) {
trace_amdgpu_vm_grab_id(vm_id->id, ring->idx);
return 0;
if (vm_id->id) {
unsigned id = vm_id->id;
long owner;
owner = atomic_long_read(&adev->vm_manager.ids[id].owner);
if (owner == (long)vm) {
trace_amdgpu_vm_grab_id(vm_id->id, ring->idx);
return 0;
}
}
/* we definately need to flush */
@ -154,7 +159,7 @@ int amdgpu_vm_grab_id(struct amdgpu_vm *vm, struct amdgpu_ring *ring,
/* skip over VMID 0, since it is the system VM */
for (i = 1; i < adev->vm_manager.nvm; ++i) {
struct fence *fence = adev->vm_manager.active[i];
struct fence *fence = adev->vm_manager.ids[i].active;
struct amdgpu_ring *fring;
if (fence == NULL) {
@ -176,7 +181,7 @@ int amdgpu_vm_grab_id(struct amdgpu_vm *vm, struct amdgpu_ring *ring,
if (choices[i]) {
struct fence *fence;
fence = adev->vm_manager.active[choices[i]];
fence = adev->vm_manager.ids[choices[i]].active;
vm_id->id = choices[i];
trace_amdgpu_vm_grab_id(choices[i], ring->idx);
@ -207,24 +212,21 @@ void amdgpu_vm_flush(struct amdgpu_ring *ring,
uint64_t pd_addr = amdgpu_bo_gpu_offset(vm->page_directory);
struct amdgpu_vm_id *vm_id = &vm->ids[ring->idx];
struct fence *flushed_updates = vm_id->flushed_updates;
bool is_earlier = false;
bool is_later;
if (flushed_updates && updates) {
BUG_ON(flushed_updates->context != updates->context);
is_earlier = (updates->seqno - flushed_updates->seqno <=
INT_MAX) ? true : false;
}
if (pd_addr != vm_id->pd_gpu_addr || !flushed_updates ||
is_earlier) {
if (!flushed_updates)
is_later = true;
else if (!updates)
is_later = false;
else
is_later = fence_is_later(updates, flushed_updates);
if (pd_addr != vm_id->pd_gpu_addr || is_later) {
trace_amdgpu_vm_flush(pd_addr, ring->idx, vm_id->id);
if (is_earlier) {
if (is_later) {
vm_id->flushed_updates = fence_get(updates);
fence_put(flushed_updates);
}
if (!flushed_updates)
vm_id->flushed_updates = fence_get(updates);
vm_id->pd_gpu_addr = pd_addr;
amdgpu_ring_emit_vm_flush(ring, vm_id->id, vm_id->pd_gpu_addr);
}
@ -244,16 +246,14 @@ void amdgpu_vm_flush(struct amdgpu_ring *ring,
*/
void amdgpu_vm_fence(struct amdgpu_device *adev,
struct amdgpu_vm *vm,
struct amdgpu_fence *fence)
struct fence *fence)
{
unsigned ridx = fence->ring->idx;
unsigned vm_id = vm->ids[ridx].id;
struct amdgpu_ring *ring = amdgpu_ring_from_fence(fence);
unsigned vm_id = vm->ids[ring->idx].id;
fence_put(adev->vm_manager.active[vm_id]);
adev->vm_manager.active[vm_id] = fence_get(&fence->base);
fence_put(vm->ids[ridx].last_id_use);
vm->ids[ridx].last_id_use = fence_get(&fence->base);
fence_put(adev->vm_manager.ids[vm_id].active);
adev->vm_manager.ids[vm_id].active = fence_get(fence);
atomic_long_set(&adev->vm_manager.ids[vm_id].owner, (long)vm);
}
/**
@ -332,6 +332,8 @@ int amdgpu_vm_free_job(struct amdgpu_job *job)
*
* @adev: amdgpu_device pointer
* @bo: bo to clear
*
* need to reserve bo first before calling it.
*/
static int amdgpu_vm_clear_bo(struct amdgpu_device *adev,
struct amdgpu_bo *bo)
@ -343,24 +345,20 @@ static int amdgpu_vm_clear_bo(struct amdgpu_device *adev,
uint64_t addr;
int r;
r = amdgpu_bo_reserve(bo, false);
if (r)
return r;
r = reservation_object_reserve_shared(bo->tbo.resv);
if (r)
return r;
r = ttm_bo_validate(&bo->tbo, &bo->placement, true, false);
if (r)
goto error_unreserve;
goto error;
addr = amdgpu_bo_gpu_offset(bo);
entries = amdgpu_bo_size(bo) / 8;
ib = kzalloc(sizeof(struct amdgpu_ib), GFP_KERNEL);
if (!ib)
goto error_unreserve;
goto error;
r = amdgpu_ib_get(ring, NULL, entries * 2 + 64, ib);
if (r)
@ -378,16 +376,14 @@ static int amdgpu_vm_clear_bo(struct amdgpu_device *adev,
if (!r)
amdgpu_bo_fence(bo, fence, true);
fence_put(fence);
if (amdgpu_enable_scheduler) {
amdgpu_bo_unreserve(bo);
if (amdgpu_enable_scheduler)
return 0;
}
error_free:
amdgpu_ib_free(adev, ib);
kfree(ib);
error_unreserve:
amdgpu_bo_unreserve(bo);
error:
return r;
}
@ -989,7 +985,7 @@ struct amdgpu_bo_va *amdgpu_vm_bo_add(struct amdgpu_device *adev,
* Add a mapping of the BO at the specefied addr into the VM.
* Returns 0 for success, error for failure.
*
* Object has to be reserved and gets unreserved by this function!
* Object has to be reserved and unreserved outside!
*/
int amdgpu_vm_bo_map(struct amdgpu_device *adev,
struct amdgpu_bo_va *bo_va,
@ -1005,30 +1001,27 @@ int amdgpu_vm_bo_map(struct amdgpu_device *adev,
/* validate the parameters */
if (saddr & AMDGPU_GPU_PAGE_MASK || offset & AMDGPU_GPU_PAGE_MASK ||
size == 0 || size & AMDGPU_GPU_PAGE_MASK) {
amdgpu_bo_unreserve(bo_va->bo);
size == 0 || size & AMDGPU_GPU_PAGE_MASK)
return -EINVAL;
}
/* make sure object fit at this offset */
eaddr = saddr + size;
if ((saddr >= eaddr) || (offset + size > amdgpu_bo_size(bo_va->bo))) {
amdgpu_bo_unreserve(bo_va->bo);
if ((saddr >= eaddr) || (offset + size > amdgpu_bo_size(bo_va->bo)))
return -EINVAL;
}
last_pfn = eaddr / AMDGPU_GPU_PAGE_SIZE;
if (last_pfn > adev->vm_manager.max_pfn) {
dev_err(adev->dev, "va above limit (0x%08X > 0x%08X)\n",
last_pfn, adev->vm_manager.max_pfn);
amdgpu_bo_unreserve(bo_va->bo);
return -EINVAL;
}
saddr /= AMDGPU_GPU_PAGE_SIZE;
eaddr /= AMDGPU_GPU_PAGE_SIZE;
spin_lock(&vm->it_lock);
it = interval_tree_iter_first(&vm->va, saddr, eaddr - 1);
spin_unlock(&vm->it_lock);
if (it) {
struct amdgpu_bo_va_mapping *tmp;
tmp = container_of(it, struct amdgpu_bo_va_mapping, it);
@ -1036,14 +1029,12 @@ int amdgpu_vm_bo_map(struct amdgpu_device *adev,
dev_err(adev->dev, "bo %p va 0x%010Lx-0x%010Lx conflict with "
"0x%010lx-0x%010lx\n", bo_va->bo, saddr, eaddr,
tmp->it.start, tmp->it.last + 1);
amdgpu_bo_unreserve(bo_va->bo);
r = -EINVAL;
goto error;
}
mapping = kmalloc(sizeof(*mapping), GFP_KERNEL);
if (!mapping) {
amdgpu_bo_unreserve(bo_va->bo);
r = -ENOMEM;
goto error;
}
@ -1055,7 +1046,9 @@ int amdgpu_vm_bo_map(struct amdgpu_device *adev,
mapping->flags = flags;
list_add(&mapping->list, &bo_va->invalids);
spin_lock(&vm->it_lock);
interval_tree_insert(&mapping->it, &vm->va);
spin_unlock(&vm->it_lock);
trace_amdgpu_vm_bo_map(bo_va, mapping);
/* Make sure the page tables are allocated */
@ -1067,8 +1060,6 @@ int amdgpu_vm_bo_map(struct amdgpu_device *adev,
if (eaddr > vm->max_pde_used)
vm->max_pde_used = eaddr;
amdgpu_bo_unreserve(bo_va->bo);
/* walk over the address space and allocate the page tables */
for (pt_idx = saddr; pt_idx <= eaddr; ++pt_idx) {
struct reservation_object *resv = vm->page_directory->tbo.resv;
@ -1077,13 +1068,11 @@ int amdgpu_vm_bo_map(struct amdgpu_device *adev,
if (vm->page_tables[pt_idx].bo)
continue;
ww_mutex_lock(&resv->lock, NULL);
r = amdgpu_bo_create(adev, AMDGPU_VM_PTE_COUNT * 8,
AMDGPU_GPU_PAGE_SIZE, true,
AMDGPU_GEM_DOMAIN_VRAM,
AMDGPU_GEM_CREATE_NO_CPU_ACCESS,
NULL, resv, &pt);
ww_mutex_unlock(&resv->lock);
if (r)
goto error_free;
@ -1101,7 +1090,9 @@ int amdgpu_vm_bo_map(struct amdgpu_device *adev,
error_free:
list_del(&mapping->list);
spin_lock(&vm->it_lock);
interval_tree_remove(&mapping->it, &vm->va);
spin_unlock(&vm->it_lock);
trace_amdgpu_vm_bo_unmap(bo_va, mapping);
kfree(mapping);
@ -1119,7 +1110,7 @@ error:
* Remove a mapping of the BO at the specefied addr from the VM.
* Returns 0 for success, error for failure.
*
* Object has to be reserved and gets unreserved by this function!
* Object has to be reserved and unreserved outside!
*/
int amdgpu_vm_bo_unmap(struct amdgpu_device *adev,
struct amdgpu_bo_va *bo_va,
@ -1144,21 +1135,20 @@ int amdgpu_vm_bo_unmap(struct amdgpu_device *adev,
break;
}
if (&mapping->list == &bo_va->invalids) {
amdgpu_bo_unreserve(bo_va->bo);
if (&mapping->list == &bo_va->invalids)
return -ENOENT;
}
}
list_del(&mapping->list);
spin_lock(&vm->it_lock);
interval_tree_remove(&mapping->it, &vm->va);
spin_unlock(&vm->it_lock);
trace_amdgpu_vm_bo_unmap(bo_va, mapping);
if (valid)
list_add(&mapping->list, &vm->freed);
else
kfree(mapping);
amdgpu_bo_unreserve(bo_va->bo);
return 0;
}
@ -1187,13 +1177,17 @@ void amdgpu_vm_bo_rmv(struct amdgpu_device *adev,
list_for_each_entry_safe(mapping, next, &bo_va->valids, list) {
list_del(&mapping->list);
spin_lock(&vm->it_lock);
interval_tree_remove(&mapping->it, &vm->va);
spin_unlock(&vm->it_lock);
trace_amdgpu_vm_bo_unmap(bo_va, mapping);
list_add(&mapping->list, &vm->freed);
}
list_for_each_entry_safe(mapping, next, &bo_va->invalids, list) {
list_del(&mapping->list);
spin_lock(&vm->it_lock);
interval_tree_remove(&mapping->it, &vm->va);
spin_unlock(&vm->it_lock);
kfree(mapping);
}
@ -1241,7 +1235,6 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm)
for (i = 0; i < AMDGPU_MAX_RINGS; ++i) {
vm->ids[i].id = 0;
vm->ids[i].flushed_updates = NULL;
vm->ids[i].last_id_use = NULL;
}
mutex_init(&vm->mutex);
vm->va = RB_ROOT;
@ -1249,7 +1242,7 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm)
INIT_LIST_HEAD(&vm->invalidated);
INIT_LIST_HEAD(&vm->cleared);
INIT_LIST_HEAD(&vm->freed);
spin_lock_init(&vm->it_lock);
pd_size = amdgpu_vm_directory_size(adev);
pd_entries = amdgpu_vm_num_pdes(adev);
@ -1269,8 +1262,14 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm)
NULL, NULL, &vm->page_directory);
if (r)
return r;
r = amdgpu_bo_reserve(vm->page_directory, false);
if (r) {
amdgpu_bo_unref(&vm->page_directory);
vm->page_directory = NULL;
return r;
}
r = amdgpu_vm_clear_bo(adev, vm->page_directory);
amdgpu_bo_unreserve(vm->page_directory);
if (r) {
amdgpu_bo_unref(&vm->page_directory);
vm->page_directory = NULL;
@ -1313,11 +1312,28 @@ void amdgpu_vm_fini(struct amdgpu_device *adev, struct amdgpu_vm *vm)
amdgpu_bo_unref(&vm->page_directory);
fence_put(vm->page_directory_fence);
for (i = 0; i < AMDGPU_MAX_RINGS; ++i) {
unsigned id = vm->ids[i].id;
atomic_long_cmpxchg(&adev->vm_manager.ids[id].owner,
(long)vm, 0);
fence_put(vm->ids[i].flushed_updates);
fence_put(vm->ids[i].last_id_use);
}
mutex_destroy(&vm->mutex);
}
/**
* amdgpu_vm_manager_fini - cleanup VM manager
*
* @adev: amdgpu_device pointer
*
* Cleanup the VM manager and free resources.
*/
void amdgpu_vm_manager_fini(struct amdgpu_device *adev)
{
unsigned i;
for (i = 0; i < AMDGPU_NUM_VM; ++i)
fence_put(adev->vm_manager.ids[i].active);
}

8
drivers/gpu/drm/amd/amdgpu/ci_dpm.c
View File

@ -6569,12 +6569,12 @@ static int ci_dpm_set_interrupt_state(struct amdgpu_device *adev,
switch (state) {
case AMDGPU_IRQ_STATE_DISABLE:
cg_thermal_int = RREG32_SMC(ixCG_THERMAL_INT);
cg_thermal_int &= ~CG_THERMAL_INT_CTRL__THERM_INTH_MASK_MASK;
cg_thermal_int |= CG_THERMAL_INT_CTRL__THERM_INTH_MASK_MASK;
WREG32_SMC(ixCG_THERMAL_INT, cg_thermal_int);
break;
case AMDGPU_IRQ_STATE_ENABLE:
cg_thermal_int = RREG32_SMC(ixCG_THERMAL_INT);
cg_thermal_int |= CG_THERMAL_INT_CTRL__THERM_INTH_MASK_MASK;
cg_thermal_int &= ~CG_THERMAL_INT_CTRL__THERM_INTH_MASK_MASK;
WREG32_SMC(ixCG_THERMAL_INT, cg_thermal_int);
break;
default:
@ -6586,12 +6586,12 @@ static int ci_dpm_set_interrupt_state(struct amdgpu_device *adev,
switch (state) {
case AMDGPU_IRQ_STATE_DISABLE:
cg_thermal_int = RREG32_SMC(ixCG_THERMAL_INT);
cg_thermal_int &= ~CG_THERMAL_INT_CTRL__THERM_INTL_MASK_MASK;
cg_thermal_int |= CG_THERMAL_INT_CTRL__THERM_INTL_MASK_MASK;
WREG32_SMC(ixCG_THERMAL_INT, cg_thermal_int);
break;
case AMDGPU_IRQ_STATE_ENABLE:
cg_thermal_int = RREG32_SMC(ixCG_THERMAL_INT);
cg_thermal_int |= CG_THERMAL_INT_CTRL__THERM_INTL_MASK_MASK;
cg_thermal_int &= ~CG_THERMAL_INT_CTRL__THERM_INTL_MASK_MASK;
WREG32_SMC(ixCG_THERMAL_INT, cg_thermal_int);
break;
default:

302
drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
View File

@ -268,7 +268,6 @@ static const u32 fiji_mgcg_cgcg_init[] =
mmCGTT_CP_CLK_CTRL, 0xffffffff, 0x00000100,
mmCGTT_CPC_CLK_CTRL, 0xffffffff, 0x00000100,
mmCGTT_CPF_CLK_CTRL, 0xffffffff, 0x40000100,
mmCGTT_DRM_CLK_CTRL0, 0xffffffff, 0x00600100,
mmCGTT_GDS_CLK_CTRL, 0xffffffff, 0x00000100,
mmCGTT_IA_CLK_CTRL, 0xffffffff, 0x06000100,
mmCGTT_PA_CLK_CTRL, 0xffffffff, 0x00000100,
@ -296,10 +295,6 @@ static const u32 fiji_mgcg_cgcg_init[] =
mmCGTS_SM_CTRL_REG, 0xffffffff, 0x96e00200,
mmCP_RB_WPTR_POLL_CNTL, 0xffffffff, 0x00900100,
mmRLC_CGCG_CGLS_CTRL, 0xffffffff, 0x0020003c,
mmPCIE_INDEX, 0xffffffff, 0x0140001c,
mmPCIE_DATA, 0x000f0000, 0x00000000,
mmCGTT_DRM_CLK_CTRL0, 0xff000fff, 0x00000100,
mmHDP_XDP_CGTT_BLK_CTRL, 0xc0000fff, 0x00000104,
mmCP_MEM_SLP_CNTL, 0x00000001, 0x00000001,
};
@ -1000,7 +995,7 @@ static void gfx_v8_0_gpu_early_init(struct amdgpu_device *adev)
adev->gfx.config.max_cu_per_sh = 16;
adev->gfx.config.max_sh_per_se = 1;
adev->gfx.config.max_backends_per_se = 4;
adev->gfx.config.max_texture_channel_caches = 8;
adev->gfx.config.max_texture_channel_caches = 16;
adev->gfx.config.max_gprs = 256;
adev->gfx.config.max_gs_threads = 32;
adev->gfx.config.max_hw_contexts = 8;
@ -1613,6 +1608,296 @@ static void gfx_v8_0_tiling_mode_table_init(struct amdgpu_device *adev)
WREG32(mmGB_MACROTILE_MODE0 + reg_offset, gb_tile_moden);
}
case CHIP_FIJI:
for (reg_offset = 0; reg_offset < num_tile_mode_states; reg_offset++) {
switch (reg_offset) {
case 0:
gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) |
MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING));
break;
case 1:
gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
TILE_SPLIT(ADDR_SURF_TILE_SPLIT_128B) |
MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING));
break;
case 2:
gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING));
break;
case 3:
gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
TILE_SPLIT(ADDR_SURF_TILE_SPLIT_512B) |
MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING));
break;
case 4:
gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
TILE_SPLIT(ADDR_SURF_TILE_SPLIT_2KB) |
MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING));
break;
case 5:
gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
TILE_SPLIT(ADDR_SURF_TILE_SPLIT_2KB) |
MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING));
break;
case 6:
gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
TILE_SPLIT(ADDR_SURF_TILE_SPLIT_2KB) |
MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING));
break;
case 7:
gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
PIPE_CONFIG(ADDR_SURF_P4_16x16) |
TILE_SPLIT(ADDR_SURF_TILE_SPLIT_2KB) |
MICRO_TILE_MODE_NEW(ADDR_SURF_DEPTH_MICRO_TILING));
break;
case 8:
gb_tile_moden = (ARRAY_MODE(ARRAY_LINEAR_ALIGNED) |
PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16));
break;
case 9:
gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) |
SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
break;
case 10:
gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) |
SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
break;
case 11:
gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) |
SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_8));
break;
case 12:
gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
PIPE_CONFIG(ADDR_SURF_P4_16x16) |
MICRO_TILE_MODE_NEW(ADDR_SURF_DISPLAY_MICRO_TILING) |
SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_8));
break;
case 13:
gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
break;
case 14:
gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
break;
case 15:
gb_tile_moden = (ARRAY_MODE(ARRAY_3D_TILED_THIN1) |
PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
break;
case 16:
gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_8));
break;
case 17:
gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
PIPE_CONFIG(ADDR_SURF_P4_16x16) |
MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_8));
break;
case 18:
gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THICK) |
PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_1));
break;
case 19:
gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THICK) |
PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
MICRO_TILE_MODE_NEW(ADDR_SURF_THICK_MICRO_TILING) |
SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_1));
break;
case 20:
gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THICK) |
PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
MICRO_TILE_MODE_NEW(ADDR_SURF_THICK_MICRO_TILING) |
SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_1));
break;
case 21:
gb_tile_moden = (ARRAY_MODE(ARRAY_3D_TILED_THICK) |
PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
MICRO_TILE_MODE_NEW(ADDR_SURF_THICK_MICRO_TILING) |
SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_1));
break;
case 22:
gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_TILED_THICK) |
PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
MICRO_TILE_MODE_NEW(ADDR_SURF_THICK_MICRO_TILING) |
SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_1));
break;
case 23:
gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_TILED_THICK) |
PIPE_CONFIG(ADDR_SURF_P4_16x16) |
MICRO_TILE_MODE_NEW(ADDR_SURF_THICK_MICRO_TILING) |
SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_1));
break;
case 24:
gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THICK) |
PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
MICRO_TILE_MODE_NEW(ADDR_SURF_THIN_MICRO_TILING) |
SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_1));
break;
case 25:
gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_XTHICK) |
PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
MICRO_TILE_MODE_NEW(ADDR_SURF_THICK_MICRO_TILING) |
SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_1));
break;
case 26:
gb_tile_moden = (ARRAY_MODE(ARRAY_3D_TILED_XTHICK) |
PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
MICRO_TILE_MODE_NEW(ADDR_SURF_THICK_MICRO_TILING) |
SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_1));
break;
case 27:
gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) |
SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
break;
case 28:
gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) |
SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_2));
break;
case 29:
gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) |
SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_8));
break;
case 30:
gb_tile_moden = (ARRAY_MODE(ARRAY_PRT_TILED_THIN1) |
PIPE_CONFIG(ADDR_SURF_P4_16x16) |
MICRO_TILE_MODE_NEW(ADDR_SURF_ROTATED_MICRO_TILING) |
SAMPLE_SPLIT(ADDR_SURF_SAMPLE_SPLIT_8));
break;
default:
gb_tile_moden = 0;
break;
}
adev->gfx.config.tile_mode_array[reg_offset] = gb_tile_moden;
WREG32(mmGB_TILE_MODE0 + reg_offset, gb_tile_moden);
}
for (reg_offset = 0; reg_offset < num_secondary_tile_mode_states; reg_offset++) {
switch (reg_offset) {
case 0:
gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
NUM_BANKS(ADDR_SURF_8_BANK));
break;
case 1:
gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
NUM_BANKS(ADDR_SURF_8_BANK));
break;
case 2:
gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
NUM_BANKS(ADDR_SURF_8_BANK));
break;
case 3:
gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
NUM_BANKS(ADDR_SURF_8_BANK));
break;
case 4:
gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
NUM_BANKS(ADDR_SURF_8_BANK));
break;
case 5:
gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
NUM_BANKS(ADDR_SURF_8_BANK));
break;
case 6:
gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
NUM_BANKS(ADDR_SURF_8_BANK));
break;
case 8:
gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_8) |
MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
NUM_BANKS(ADDR_SURF_8_BANK));
break;
case 9:
gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
NUM_BANKS(ADDR_SURF_8_BANK));
break;
case 10:
gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
NUM_BANKS(ADDR_SURF_8_BANK));
break;
case 11:
gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
NUM_BANKS(ADDR_SURF_8_BANK));
break;
case 12:
gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
NUM_BANKS(ADDR_SURF_8_BANK));
break;
case 13:
gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2) |
NUM_BANKS(ADDR_SURF_8_BANK));
break;
case 14:
gb_tile_moden = (BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1) |
NUM_BANKS(ADDR_SURF_4_BANK));
break;
case 7:
/* unused idx */
continue;
default:
gb_tile_moden = 0;
break;
}
adev->gfx.config.macrotile_mode_array[reg_offset] = gb_tile_moden;
WREG32(mmGB_MACROTILE_MODE0 + reg_offset, gb_tile_moden);
}
break;
case CHIP_TONGA:
for (reg_offset = 0; reg_offset < num_tile_mode_states; reg_offset++) {
switch (reg_offset) {
@ -2971,10 +3256,13 @@ static int gfx_v8_0_cp_gfx_start(struct amdgpu_device *adev)
amdgpu_ring_write(ring, mmPA_SC_RASTER_CONFIG - PACKET3_SET_CONTEXT_REG_START);
switch (adev->asic_type) {
case CHIP_TONGA:
case CHIP_FIJI:
amdgpu_ring_write(ring, 0x16000012);
amdgpu_ring_write(ring, 0x0000002A);
break;
case CHIP_FIJI:
amdgpu_ring_write(ring, 0x3a00161a);
amdgpu_ring_write(ring, 0x0000002e);
break;
case CHIP_TOPAZ:
case CHIP_CARRIZO:
amdgpu_ring_write(ring, 0x00000002);

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