Hi Takahiro, On Thu, 2015-03-26 at 17:28 +0900, AKASHI Takahiro wrote: > On system kernel, the memory region used by crash dump kernel must be > specified by "crashkernel=X at Y" boot parameter. reserve_crashkernel() > will allocate the region in "System RAM" and reserve it for later use. > > On crash dump kernel, memory region information in system kernel is > described in a specific region specified by "elfcorehdr=X at Y" boot parameter. > reserve_elfcorehdr() will set aside the region to avoid data destruction > by the kernel. > > Crash dump kernel will access memory regions in system kernel via > copy_oldmem_page(), which reads a page with ioremap'ing it assuming that s/page with/page by/ > such pages are not part of main memory of crash dump kernel. > This is true under non-UEFI environment because kexec-tools modifies > a device tree adding "usablemem" attributes to memory sections. > Under UEFI, however, this is not true because UEFI remove memory sections > in a device tree and export all the memory regions, even though they belong > to system kernel. > > So we should add "mem=X[MG]" boot parameter to limit the meory size and s/meory/memory/ > avoid hitting the following assertion in ioremap(): > if (WARN_ON(pfn_valid(__phys_to_pfn(phys_addr)))) > return NULL; > > Signed-off-by: AKASHI Takahiro <takahiro.akashi at linaro.org> > --- > arch/arm64/kernel/Makefile | 1 + > arch/arm64/kernel/crash_dump.c | 71 ++++++++++++++++++++++++++++++++++++ > arch/arm64/kernel/setup.c | 78 ++++++++++++++++++++++++++++++++++++++++ > 3 files changed, 150 insertions(+) > create mode 100644 arch/arm64/kernel/crash_dump.c > > diff --git a/arch/arm64/kernel/Makefile b/arch/arm64/kernel/Makefile > index da9a7ee..3c24d4e 100644 > --- a/arch/arm64/kernel/Makefile > +++ b/arch/arm64/kernel/Makefile > @@ -36,6 +36,7 @@ arm64-obj-$(CONFIG_EFI) += efi.o efi-stub.o efi-entry.o > arm64-obj-$(CONFIG_PCI) += pci.o > arm64-obj-$(CONFIG_ARMV8_DEPRECATED) += armv8_deprecated.o > arm64-obj-$(CONFIG_KEXEC) += machine_kexec.o relocate_kernel.o > +arm64-obj-$(CONFIG_CRASH_DUMP) += crash_dump.o > > obj-y += $(arm64-obj-y) vdso/ > obj-m += $(arm64-obj-m) > diff --git a/arch/arm64/kernel/crash_dump.c b/arch/arm64/kernel/crash_dump.c > new file mode 100644 > index 0000000..dd31b2e > --- /dev/null > +++ b/arch/arm64/kernel/crash_dump.c > @@ -0,0 +1,71 @@ > +/* > + * arch/arm64/kernel/crash_dump.c I would recommend against adding paths in source files. It often happens that files with paths are moved, but the file comments are not updated. > + * > + * Copyright (C) 2014 Linaro Limited > + * Author: AKASHI Takahiro <takahiro.akashi at linaro.org> > + * > + * This program is free software; you can redistribute it and/or modify > + * it under the terms of the GNU General Public License version 2 as > + * published by the Free Software Foundation. > + */ > + > +#include <linux/crash_dump.h> > +#include <linux/errno.h> > +#include <linux/io.h> > +#include <linux/memblock.h> > +#include <linux/uaccess.h> > +#include <asm/memory.h> > + > +/** > + * copy_oldmem_page() - copy one page from old kernel memory > + * @pfn: page frame number to be copied > + * @buf: buffer where the copied page is placed > + * @csize: number of bytes to copy > + * @offset: offset in bytes into the page > + * @userbuf: if set, @buf is int he user address space s/is int he user/is a user/ > + * > + * This function copies one page from old kernel memory into buffer pointed by > + * @buf. If @buf is in userspace, set @userbuf to %1. Returns number of bytes > + * copied or negative error in case of failure. > + */ > +ssize_t copy_oldmem_page(unsigned long pfn, char *buf, > + size_t csize, unsigned long offset, > + int userbuf) Since userbuf is a binary flag, I think type bool would be better. Change the comments from 'set' and '1' to 'true'. Should offset be type size_t? > +{ > + void *vaddr; > + > + if (!csize) > + return 0; > + > + vaddr = ioremap(pfn << PAGE_SHIFT, PAGE_SIZE); > + if (!vaddr) > + return -ENOMEM; > + > + if (userbuf) { > + if (copy_to_user(buf, vaddr + offset, csize)) { > + iounmap(vaddr); > + return -EFAULT; > + } > + } else { > + memcpy(buf, vaddr + offset, csize); > + } > + > + iounmap(vaddr); > + > + return csize; > +} > + > +/** > + * elfcorehdr_read - read from ELF core header > + * @buf: buffer where the data is placed > + * @csize: number of bytes to read > + * @ppos: address in the memory > + * > + * This function reads @count bytes from elf core header which exists > + * on crash dump kernel's memory. > + */ > +ssize_t elfcorehdr_read(char *buf, size_t count, u64 *ppos) Should ppos be type phys_addr_t *? > +{ > + memcpy(buf, phys_to_virt((phys_addr_t)*ppos), count); > + return count; > +} > diff --git a/arch/arm64/kernel/setup.c b/arch/arm64/kernel/setup.c > index e8420f6..daaed93 100644 > --- a/arch/arm64/kernel/setup.c > +++ b/arch/arm64/kernel/setup.c > @@ -323,6 +323,69 @@ static void __init setup_machine_fdt(phys_addr_t dt_phys) > dump_stack_set_arch_desc("%s (DT)", of_flat_dt_get_machine_name()); > } > > +#ifdef CONFIG_KEXEC > +/* > + * reserve_crashkernel() - reserves memory for crash kernel > + * > + * This function reserves memory area given in "crashkernel=" kernel command > + * line parameter. The memory reserved is used by a dump capture kernel when > + * primary kernel is crashing. > + */ > +static void __init reserve_crashkernel(void) > +{ > + unsigned long long crash_size, crash_base; > + int ret; > + > + /* use ULONG_MAX since we don't know system memory size here. */ > + ret = parse_crashkernel(boot_command_line, ULONG_MAX, > + &crash_size, &crash_base); > + if (ret) > + return; > + > + ret = memblock_reserve(crash_base, crash_size); > + if (ret < 0) { > + pr_warn("crashkernel reservation failed - memory is in use (0x%lx)\n", > + (unsigned long)crash_base); Can we use 0x%llx here and below and avoid these casts? > + return; > + } > + > + pr_info("Reserving %ldMB of memory at %ldMB for crashkernel\n", > + (unsigned long)(crash_size >> 20), > + (unsigned long)(crash_base >> 20)); > + > + crashk_res.start = crash_base; > + crashk_res.end = crash_base + crash_size - 1; > +} > +#endif /* CONFIG_KEXEC */ > + > +#ifdef CONFIG_CRASH_DUMP > +/* > + * reserve_elfcorehdr() - reserves memory for elf core header > + * > + * This function reserves memory area given in "elfcorehdr=" kernel command > + * line parameter. The memory reserved is used by a dump capture kernel to > + * identify the memory used by primary kernel. > + */ > +static void __init reserve_elfcorehdr(void) > +{ > + int ret; > + > + if (!elfcorehdr_size) > + return; > + > + ret = memblock_reserve(elfcorehdr_addr, elfcorehdr_size); > + if (ret < 0) { > + pr_warn("elfcorehdr reservation failed - memory is in use (0x%lx)\n", > + (unsigned long)elfcorehdr_addr); > + return; > + } > + > + pr_info("Reserving %ldKB of memory at %ldMB for elfcorehdr\n", > + (unsigned long)(elfcorehdr_size >> 10), > + (unsigned long)(elfcorehdr_addr >> 20)); > +} > +#endif /* CONFIG_CRASH_DUMP */ > + > static void __init request_standard_resources(void) > { > struct memblock_region *region; > @@ -378,10 +441,25 @@ void __init setup_arch(char **cmdline_p) > local_async_enable(); > > efi_init(); Maybe have a blank line here? > + /* > + * reserve_crashkernel() and reserver_elfcorehdr() must be called s/reserver_elfcorehdr/reserve_elfcorehdr/ > + * before arm64_bootmem_init() because dma_contiguous_reserve() > + * may conflict with those regions. > + */ > +#ifdef CONFIG_KEXEC > + reserve_crashkernel(); > +#endif > +#ifdef CONFIG_CRASH_DUMP > + reserve_elfcorehdr(); > +#endif > arm64_memblock_init(); > > paging_init(); > request_standard_resources(); > +#ifdef CONFIG_KEXEC > + /* kexec-tool will detect the region with /proc/iomem */ There are more kexec user programs than kexec-tools, so I think something like 'user space tools' would be better. > + insert_resource(&iomem_resource, &crashk_res); > +#endif > > early_ioremap_reset(); >