Patch series "sections: Unify kernel sections range check and use", v4.
There are three head files(kallsyms.h, kernel.h and sections.h) which
include the kernel sections range check, let's make some cleanup and unify
them.
1. cleanup arch specific text/data check and fix address boundary check
in kallsyms.h
2. make all the basic/core kernel range check function into sections.h
3. update all the callers, and use the helper in sections.h to simplify
the code
After this series, we have 5 APIs about kernel sections range check in
sections.h
* is_kernel_rodata() --- already in sections.h
* is_kernel_core_data() --- come from core_kernel_data() in kernel.h
* is_kernel_inittext() --- come from kernel.h and kallsyms.h
* __is_kernel_text() --- add new internal helper
* __is_kernel() --- add new internal helper
Note: For the last two helpers, people should not use directly, consider to
use corresponding function in kallsyms.h.
This patch (of 11):
Remove arch specific text and data check after commit 4ba66a9760 ("arch:
remove blackfin port"), no need arch-specific text/data check.
Link: https://lkml.kernel.org/r/[email protected]
Link: https://lkml.kernel.org/r/[email protected]
Signed-off-by: Kefeng Wang <[email protected]>
Reviewed-by: Sergey Senozhatsky <[email protected]>
Cc: Arnd Bergmann <[email protected]>
Cc: Steven Rostedt <[email protected]>
Cc: Ingo Molnar <[email protected]>
Cc: David S. Miller <[email protected]>
Cc: Alexei Starovoitov <[email protected]>
Cc: Andrey Ryabinin <[email protected]>
Cc: Michael Ellerman <[email protected]>
Cc: Benjamin Herrenschmidt <[email protected]>
Cc: Paul Mackerras <[email protected]>
Cc: Christophe Leroy <[email protected]>
Cc: Kefeng Wang <[email protected]>
Cc: Alexander Potapenko <[email protected]>
Cc: Andrey Konovalov <[email protected]>
Cc: Borislav Petkov <[email protected]>
Cc: Dmitry Vyukov <[email protected]>
Cc: Ivan Kokshaysky <[email protected]>
Cc: Matt Turner <[email protected]>
Cc: Michal Simek <[email protected]>
Cc: Petr Mladek <[email protected]>
Cc: Richard Henderson <[email protected]>
Cc: Thomas Gleixner <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Percpu embedded first chunk allocator is the firstly option, but it
could fails on ARM64, eg,
percpu: max_distance=0x5fcfdc640000 too large for vmalloc space 0x781fefff0000
percpu: max_distance=0x600000540000 too large for vmalloc space 0x7dffb7ff0000
percpu: max_distance=0x5fff9adb0000 too large for vmalloc space 0x5dffb7ff0000
then we could get
WARNING: CPU: 15 PID: 461 at vmalloc.c:3087 pcpu_get_vm_areas+0x488/0x838
and the system could not boot successfully.
Let's implement page mapping percpu first chunk allocator as a fallback
to the embedding allocator to increase the robustness of the system.
Link: https://lkml.kernel.org/r/[email protected]
Signed-off-by: Kefeng Wang <[email protected]>
Reviewed-by: Catalin Marinas <[email protected]>
Cc: Andrey Konovalov <[email protected]>
Cc: Andrey Ryabinin <[email protected]>
Cc: Dmitry Vyukov <[email protected]>
Cc: Greg Kroah-Hartman <[email protected]>
Cc: Marco Elver <[email protected]>
Cc: Will Deacon <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
Percpu embedded first chunk allocator is the firstly option, but it
could fail on ARM64, eg,
percpu: max_distance=0x5fcfdc640000 too large for vmalloc space 0x781fefff0000
percpu: max_distance=0x600000540000 too large for vmalloc space 0x7dffb7ff0000
percpu: max_distance=0x5fff9adb0000 too large for vmalloc space 0x5dffb7ff0000
then we could get to
WARNING: CPU: 15 PID: 461 at vmalloc.c:3087 pcpu_get_vm_areas+0x488/0x838
and the system cannot boot successfully.
Let's implement page mapping percpu first chunk allocator as a fallback
to the embedding allocator to increase the robustness of the system.
Also fix a crash when both NEED_PER_CPU_PAGE_FIRST_CHUNK and
KASAN_VMALLOC enabled.
Tested on ARM64 qemu with cmdline "percpu_alloc=page".
This patch (of 3):
There are some fixed locations in the vmalloc area be reserved in
ARM(see iotable_init()) and ARM64(see map_kernel()), but for
pcpu_page_first_chunk(), it calls vm_area_register_early() and choose
VMALLOC_START as the start address of vmap area which could be
conflicted with above address, then could trigger a BUG_ON in
vm_area_add_early().
Let's choose a suit start address by traversing the vmlist.
Link: https://lkml.kernel.org/r/[email protected]
Link: https://lkml.kernel.org/r/[email protected]
Signed-off-by: Kefeng Wang <[email protected]>
Reviewed-by: Catalin Marinas <[email protected]>
Cc: Will Deacon <[email protected]>
Cc: Andrey Ryabinin <[email protected]>
Cc: Andrey Konovalov <[email protected]>
Cc: Dmitry Vyukov <[email protected]>
Cc: Marco Elver <[email protected]>
Cc: Greg Kroah-Hartman <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
After commit f227f0faf6 ("slub: fix unreclaimable slab stat for bulk
free"), the check for free nonslab page is replaced by VM_BUG_ON_PAGE,
which only check with CONFIG_DEBUG_VM enabled, but this config may
impact performance, so it only for debug.
Commit 0937502af7 ("slub: Add check for kfree() of non slab objects.")
add the ability, which should be needed in any configs to catch the
invalid free, they even could be potential issue, eg, memory corruption,
use after free and double free, so replace VM_BUG_ON_PAGE to
WARN_ON_ONCE, add object address printing to help use to debug the
issue.
Link: https://lkml.kernel.org/r/[email protected]
Signed-off-by: Kefeng Wang <[email protected]>
Cc: Matthew Wilcox <[email protected]>
Cc: Shakeel Butt <[email protected]>
Cc: Vlastimil Babka <[email protected]>
Cc: Christoph Lameter <[email protected]>
Cc: Pekka Enberg <[email protected]>
Cc: David Rienjes <[email protected]>
Cc: Joonsoo Kim <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
It seems that something is wrong when patch "riscv/vdso:
Refactor asm/vdso.h" is merged.
Let's fix the merge issue.
Fixes: 8edab02386 ("Merge remote-tracking branch 'palmer/riscv-vdso-cleanup' into for-next")
Signed-off-by: Kefeng Wang <[email protected]>
Signed-off-by: Palmer Dabbelt <[email protected]>
When user code execution with privilege mode, it will lead to
infinite loop in the page fault handler if ARM_LPAE enabled,
The issue could be reproduced with
"echo EXEC_USERSPACE > /sys/kernel/debug/provoke-crash/DIRECT"
As Permission fault shows in ARM spec,
IFSR format when using the Short-descriptor translation table format
Permission fault: 01101 First level 01111 Second level
IFSR format when using the Long-descriptor translation table format
Permission fault: 0011LL LL bits indicate levelb.
Add is_permission_fault() function to check permission fault and die
if permission fault occurred under instruction fault in do_page_fault().
Fixes: 1d4d37159d ("ARM: 8235/1: Support for the PXN CPU feature on ARMv7")
Signed-off-by: Kefeng Wang <[email protected]>
Signed-off-by: Russell King (Oracle) <[email protected]>
Provide die_kernel_fault() helper to do the kernel fault reporting,
which with msg argument, it could report different message in different
scenes, and the later patch "ARM: mm: Fix PXN process with LPAE feature"
will use it.
Signed-off-by: Kefeng Wang <[email protected]>
Signed-off-by: Russell King (Oracle) <[email protected]>
Now the show_pts() will dump the virtual (hashed) address of page
table base, it is useless, kill it.
Signed-off-by: Kefeng Wang <[email protected]>
Signed-off-by: Russell King (Oracle) <[email protected]>
Now the write fault check in do_page_fault() and access_error() twice,
we can cleanup access_error(), and make the fault check and vma flags set
into do_page_fault() directly, then pass the vma flags to __do_page_fault.
No functional change.
Signed-off-by: Kefeng Wang <[email protected]>
Signed-off-by: Russell King (Oracle) <[email protected]>
The __do_page_fault() won't use task_struct argument, kill it
and also use current->mm directly in do_page_fault().
No functional change.
Signed-off-by: Kefeng Wang <[email protected]>
Signed-off-by: Russell King (Oracle) <[email protected]>
Clean up the multiple goto statements and drops local variable
vm_fault_t fault, which will make the __do_page_fault() much more
readability.
No functional change.
Signed-off-by: Kefeng Wang <[email protected]>
Signed-off-by: Russell King (Oracle) <[email protected]>
of_amba_device_create() uses irq_of_parse_and_map() to translate
a DT interrupt specification into a Linux virtual interrupt number.
But it doesn't properly handle the case where the interrupt controller
is not yet available, eg, when pl011 interrupt is connected to MBIGEN
interrupt controller, because the mbigen initialization is too late,
which will lead to no IRQ due to no IRQ domain found, log is shown below,
"irq: no irq domain found for uart0 !"
use of_irq_get() to return -EPROBE_DEFER as above, and in the function
amba_device_try_add()/amba_device_add(), it will properly handle in such
case, also return 0 in other fail cases to be consistent as before.
Cc: Rob Herring <[email protected]>
Cc: Frank Rowand <[email protected]>
Reported-by: Ruizhe Lin <[email protected]>
Reviewed-by: Rob Herring <[email protected]>
Signed-off-by: Kefeng Wang <[email protected]>
Signed-off-by: Russell King (Oracle) <[email protected]>
After commit 77a7300aba ("of/irq: Get rid of NO_IRQ usage"),
no irq case has been removed, irq_of_parse_and_map() will return
0 in all cases when get error from parse and map an interrupt into
linux virq space.
amba_device_register() is only used on no-DT initialization, see
s3c64xx_pl080_init() arch/arm/mach-s3c/pl080.c
ep93xx_init_devices() arch/arm/mach-ep93xx/core.c
They won't set -1 to irq[0], so no need the warn.
This reverts commit 2eac58d502.
Reviewed-by: Rob Herring <[email protected]>
Signed-off-by: Kefeng Wang <[email protected]>
Signed-off-by: Russell King (Oracle) <[email protected]>
No one use the following functions, kill them.
amba_aphb_device_add()
amba_apb_device_add()
amba_apb_device_add_res()
amba_ahb_device_add()
amba_ahb_device_add_res()
Cc: Linus Walleij <[email protected]>
Reviewed-by: Rob Herring <[email protected]>
Signed-off-by: Kefeng Wang <[email protected]>
Signed-off-by: Russell King (Oracle) <[email protected]>
This enlarges the bits availiable for stack randomisation on RV64 from
the default of 8MiB to 1GiB, to match arm64 and x86.
Also, update the documentation to reflect our support for stack
randomisation.
Signed-off-by: Kefeng Wang <[email protected]>
[Palmer: commit text]
Signed-off-by: Palmer Dabbelt <[email protected]>
This patch adds support to allocate gigantic hugepages using CMA by
specifying the hugetlb_cma= kernel parameter. This is only supported on
RV64.
Reviewed-by: Alexandre Ghiti <[email protected]>
Signed-off-by: Kefeng Wang <[email protected]>
Signed-off-by: Palmer Dabbelt <[email protected]>
There is already a memory_intersects() helper in sections.h,
use memory_intersects() directly instead of private overlap().
Signed-off-by: Kefeng Wang <[email protected]>
Signed-off-by: Christoph Hellwig <[email protected]>
DO_ONCE
DEFINE_STATIC_KEY_TRUE(___once_key);
__do_once_done
once_disable_jump(once_key);
INIT_WORK(&w->work, once_deferred);
struct once_work *w;
w->key = key;
schedule_work(&w->work); module unload
//*the key is
destroy*
process_one_work
once_deferred
BUG_ON(!static_key_enabled(work->key));
static_key_count((struct static_key *)x) //*access key, crash*
When module uses DO_ONCE mechanism, it could crash due to the above
concurrency problem, we could reproduce it with link[1].
Fix it by add/put module refcount in the once work process.
[1] https://lore.kernel.org/netdev/[email protected]/
Cc: Hannes Frederic Sowa <[email protected]>
Cc: Daniel Borkmann <[email protected]>
Cc: David S. Miller <[email protected]>
Cc: Eric Dumazet <[email protected]>
Reported-by: Minmin chen <[email protected]>
Signed-off-by: Kefeng Wang <[email protected]>
Acked-by: Hannes Frederic Sowa <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
The interrupt controllers on riscv support both edge and level triggered
interrupts, it's useful to provide that information in /proc/interrupts.
Signed-off-by: Kefeng Wang <[email protected]>
Signed-off-by: Palmer Dabbelt <[email protected]>
Enable generic idle loop to support for hlt/nohlt command line options
to override default idle loop behavior.
Signed-off-by: Kefeng Wang <[email protected]>
Signed-off-by: Palmer Dabbelt <[email protected]>
The timer interrupt and the perf interrupt on riscv are with
IRQF_PERCPU, so it's safe to allow forced interrupt threading.
Signed-off-by: Kefeng Wang <[email protected]>
Signed-off-by: Palmer Dabbelt <[email protected]>
The memblock_enforce_memory_limit() could change the memblock
range, so move the dram_end assignment after it in bootmem_init(),
then support mem= cmdline.
Signed-off-by: Kefeng Wang <[email protected]>
Signed-off-by: Palmer Dabbelt <[email protected]>
This reverts commit e220ba6022.
The VERIFY_PERMISSION_FAULT is introduced since 2009 but no
one use it, just revert it and clean unused comment.
Signed-off-by: Kefeng Wang <[email protected]>
Signed-off-by: Russell King <[email protected]>
The SWIOTLB buffer is not needed unless the physical address space
is beyond the limit of dma, only initialize swiotlb when swiotlb_force
is true or not all system memory is DMA-able.
Also move the swiotlb_init() into mem_init().
Signed-off-by: Kefeng Wang <[email protected]>
Signed-off-by: Palmer Dabbelt <[email protected]>
The _sdata/_edata is already in sections.h, drop redundant
declaration.
Also move _xiprom/_exiprom declarations at the beginning of
the file, cleanup one CONFIG_XIP_KERNEL.
Signed-off-by: Kefeng Wang <[email protected]>
Signed-off-by: Palmer Dabbelt <[email protected]>
As of commit dce4456619 ("mm/memtest: add ARCH_USE_MEMTEST"),
architectures must select ARCH_USE_MEMTESET to enable CONFIG_MEMTEST.
Signed-off-by: Kefeng Wang <[email protected]>
Fixes: f6e5aedf47 ("riscv: Add support for memtest")
Signed-off-by: Palmer Dabbelt <[email protected]>
NUMA is useless when NOMMU, and it leads some build error,
make it depend on MMU.
Reported-by: kernel test robot <[email protected]>
Signed-off-by: Kefeng Wang <[email protected]>
Signed-off-by: Palmer Dabbelt <[email protected]>
The const annotation should not be used for 'sp', or it will
become read only and lead to bad stack output.
Fixes: dec822771b ("riscv: stacktrace: Move register keyword to beginning of declaration")
Signed-off-by: Kefeng Wang <[email protected]>
Signed-off-by: Palmer Dabbelt <[email protected]>
There are two issues for RV32,
1) if use FLATMEM, it is useless to enable SPARSEMEM_STATIC.
2) if use SPARSMEM, both SPARSEMEM_VMEMMAP and SPARSEMEM_STATIC is enabled.
Fixes: d95f1a542c ("RISC-V: Implement sparsemem")
Signed-off-by: Kefeng Wang <[email protected]>
Cc: [email protected]
Signed-off-by: Palmer Dabbelt <[email protected]>
FORTIFY_SOURCE could detect various overflows at compile and run time.
ARCH_HAS_FORTIFY_SOURCE means that the architecture can be built and
run with CONFIG_FORTIFY_SOURCE. Select it in RISCV.
See more about this feature from commit 6974f0c455
("include/linux/string.h: add the option of fortified string.h functions").
Signed-off-by: Kefeng Wang <[email protected]>
Signed-off-by: Palmer Dabbelt <[email protected]>
The riscv [rv32_]defconfig enabled CONFIG_MEMTEST,
but memtest feature is not supported in RISCV.
Add early_memtest() to support for memtest.
Signed-off-by: Kefeng Wang <[email protected]>
Signed-off-by: Palmer Dabbelt <[email protected]>
After the following patches,
commit de043da0b9 ("RISC-V: Fix usage of memblock_enforce_memory_limit")
commit 1bd14a66ee ("RISC-V: Remove any memblock representing unusable memory area")
commit b10d6bca87 ("arch, drivers: replace for_each_membock() with for_each_mem_range()")
some logic is useless, kill the mem_start/start/end and unneeded code.
Reviewed-by: Atish Patra <[email protected]>
Signed-off-by: Kefeng Wang <[email protected]>
Signed-off-by: Palmer Dabbelt <[email protected]>
max_low_pfn and min_low_pfn are declared in linux/memblock.h,
and it also is included in arch/riscv/mm/init.c, drop unnecessary
declaration.
Signed-off-by: Kefeng Wang <[email protected]>
Reviewed-by: Atish Patra <[email protected]>
Signed-off-by: Palmer Dabbelt <[email protected]>
Some architectures(eg, ARM and riscv) have similar logic to
check and reserve the memory of initrd, let's provide a common
function reserve_initrd_mem() to reduce duplicated code.
Signed-off-by: Kefeng Wang <[email protected]>
Signed-off-by: Palmer Dabbelt <[email protected]>
Add the preprocessor guard in initrd.h to prevent possible
build error from the multiple inclusion of same header file
multiple time.
Signed-off-by: Kefeng Wang <[email protected]>
Signed-off-by: Palmer Dabbelt <[email protected]>
Using global sp_in_global directly to fix the following warning,
arch/riscv/kernel/stacktrace.c:31:3: warning: ‘register’ is not at beginning of declaration [-Wold-style-declaration]
31 | const register unsigned long current_sp = sp_in_global;
| ^~~~~
Signed-off-by: Kefeng Wang <[email protected]>
Signed-off-by: Palmer Dabbelt <[email protected]>
Show the function symbols of epc and ra to improve the
readability of crash reports, and align the printing
formats about the raw epc value.
Signed-off-by: Kefeng Wang <[email protected]>
Signed-off-by: Palmer Dabbelt <[email protected]>
Like commit 1149aad10b ("arm64: Add dump_backtrace() in show_regs"),
dump the stack in riscv show_regs as common code expects.
Reviewed-by: Atish Patra <[email protected]>
Signed-off-by: Kefeng Wang <[email protected]>
Signed-off-by: Palmer Dabbelt <[email protected]>
Add the machine name to kernel boot-up log, and install
the machine name to stack dump for DT boot mode.
Signed-off-by: Kefeng Wang <[email protected]>
Reviewed-by: Atish Patra <[email protected]>
Signed-off-by: Palmer Dabbelt <[email protected]>
Fix sbi_init() function declaration mismatch between RISCV_SBI
enable and disable, as it always returned 0, make it void function.
Drop some stubs which won't be used if RISCV_SBI disabled.
Signed-off-by: Kefeng Wang <[email protected]>
Reviewed-by: Atish Patra <[email protected]>
Signed-off-by: Palmer Dabbelt <[email protected]>
drivers/clocksource/timer-sp804.c:38:31: warning: symbol 'arm_sp804_timer' was not declared. Should it be static?
drivers/clocksource/timer-sp804.c:47:31: warning: symbol 'hisi_sp804_timer' was not declared. Should it be static?
drivers/clocksource/timer-sp804.c:120:12: warning: symbol 'sp804_clocksource_and_sched_clock_init' was not declared. Should it be static?
drivers/clocksource/timer-sp804.c:219:12: warning: symbol 'sp804_clockevents_init' was not declared. Should it be static?
And move __initdata after the variables.
Signed-off-by: Kefeng Wang <[email protected]>
Signed-off-by: Daniel Lezcano <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Convert to ARCH_STACKWALK to reduce duplicated code in stack trace.
Signed-off-by: Kefeng Wang <[email protected]>
Signed-off-by: Palmer Dabbelt <[email protected]>
In order to use generic arch_stack_walk() code, make stack walk
callback consistent with it.
Signed-off-by: Kefeng Wang <[email protected]>
Signed-off-by: Palmer Dabbelt <[email protected]>
RISCV_TIMER/CLINT_TIMER is required for RISC-V system, and it
provides sched_clock, which allow us to enable IRQ_TIME_ACCOUNTING.
Signed-off-by: Kefeng Wang <[email protected]>
Signed-off-by: Palmer Dabbelt <[email protected]>
riscv has selected HAVE_DMA_CONTIGUOUS, but doesn't call
dma_contiguous_reserve(). This calls dma_contiguous_reserve(), which
enables CMA.
Signed-off-by: Kefeng Wang <[email protected]>
Signed-off-by: Palmer Dabbelt <[email protected]>
Let's remove all files under riscv boot dir by using archclean rule.
Signed-off-by: Kefeng Wang <[email protected]>
Reviewed-by: Atish Patra <[email protected]>
Signed-off-by: Palmer Dabbelt <[email protected]>
make[1]: *** No rule to make target `Image.lzma'. Stop.
When make ARCH=riscv Image.lzma, it won't work, let's fix it.
Reviewed-by: Atish Patra <[email protected]>
Signed-off-by: Kefeng Wang <[email protected]>
Signed-off-by: Palmer Dabbelt <[email protected]>
Kefeng Wang