Currently, read to /proc/vmcore is done by read_oldmem() that uses ioremap/iounmap per a single page. For example, if memory is 1GB, ioremap/iounmap is called (1GB / 4KB)-times, that is, 262144 times. This causes big performance degradation due to repeated page table changes, TLB flush and build-up of VM related objects. In particular, the current main user of this mmap() is makedumpfile, which not only reads memory from /proc/vmcore but also does other processing like filtering, compression and IO work. To address the issue, this patch implements mmap() on /proc/vmcore to improve read performance. Benchmark ========= You can see two benchmarks on terabyte memory system. Both show about 40 seconds on 2TB system. This is almost equal to performance by experimental kernel-side memory filtering. - makedumpfile mmap() benchmark, by Jingbai Ma https://lkml.org/lkml/2013/3/27/19 - makedumpfile: benchmark on mmap() with /proc/vmcore on 2TB memory system https://lkml.org/lkml/2013/3/26/914 ChangeLog ========= v6 => v7) - Rebase 3.10-rc2. - Move roundup operation to note segment from patch 2/8 to patch 6/8. - Rewrite get_note_number_and_size_elf{64,32} and copy_notes_elf{64,32} in patch 6/8. v5 => v6) - Change patch order: clenaup patch => PT_LOAD change patch => vmalloc-related patch => mmap patch. - Some cleanups: improve symbol names simply, add helper functoins for processing ELF note segment and add comments for the helper functions. - Fix patch description of patch 7/8. v4 => v5) - Rebase 3.10-rc1. - Introduce remap_vmalloc_range_partial() in order to remap vmalloc memory in a part of vma area. - Allocate buffer for ELF note segment at 2nd kernel by vmalloc(). Use remap_vmalloc_range_partial() to remap the memory to userspace. v3 => v4) - Rebase 3.9-rc7. - Drop clean-up patches orthogonal to the main topic of this patch set. - Copy ELF note segments in the 2nd kernel just as in v1. Allocate vmcore objects per pages. => See [PATCH 5/8] - Map memory referenced by PT_LOAD entry directly even if the start or end of the region doesn't fit inside page boundary, no longer copy them as the previous v3. Then, holes, outside OS memory, are visible from /proc/vmcore. => See [PATCH 7/8] v2 => v3) - Rebase 3.9-rc3. - Copy program headers separately from e_phoff in ELF note segment buffer. Now there's no risk to allocate huge memory if program header table positions after memory segment. - Add cleanup patch that removes unnecessary variable. - Fix wrongly using the variable that is buffer size configurable at runtime. Instead, use the variable that has original buffer size. v1 => v2) - Clean up the existing codes: use e_phoff, and remove the assumption on PT_NOTE entries. - Fix potential bug that ELF header size is not included in exported vmcoreinfo size. - Divide patch modifying read_vmcore() into two: clean-up and primary code change. - Put ELF note segments in page-size boundary on the 1st kernel instead of copying them into the buffer on the 2nd kernel. Test ==== This patch set is composed based on v3.10-rc2, tested on x86_64, x86_32 both with 1GB and with 5GB (over 4GB) memory configurations. --- HATAYAMA Daisuke (8): vmcore: support mmap() on /proc/vmcore vmcore: calculate vmcore file size from buffer size and total size of vmcore objects vmcore: allocate ELF note segment in the 2nd kernel vmalloc memory vmalloc: introduce remap_vmalloc_range_partial vmalloc: make find_vm_area check in range vmcore: treat memory chunks referenced by PT_LOAD program header entries in page-size boundary in vmcore_list vmcore: allocate buffer for ELF headers on page-size alignment vmcore: clean up read_vmcore() fs/proc/vmcore.c | 595 +++++++++++++++++++++++++++++++++++------------ include/linux/vmalloc.h | 4 mm/vmalloc.c | 65 ++++- 3 files changed, 494 insertions(+), 170 deletions(-) -- Thanks. HATAYAMA, Daisuke
