May 2026 - Crash-utility - Crash Utility List Archives

[PATCH] crash: symbols: Optimize kernel module pseudo-symbol matching overhead

by Rui Qi

In a modern Linux environment with numerous kernel modules loaded, the crash utility must process and sort tens of thousands (often 50,000+) of module symbols. The symbol sorting function (compare_syms) and module loading loops frequently check for pseudo-symbols using macros like MODULE_PSEUDO_SYMBOL(), MODULE_START(), and others. These macros heavily rely on the STREQ()/STRNEQ() macros, which internally expand to string_exists() checks followed by a full strcmp() or strncmp(). Because these macros are evaluated millions of times during module initialization and sorting (O(N log N) complexity), the function call overhead of strcmp() becomes a noticeable performance bottleneck. By explicitly checking the first character of the symbol name (e.g., (sp)->name[0] == '_') before invoking STRNEQ(), we introduce a lightweight "early reject" mechanism. Since most regular module symbols do not start with an underscore, this short-circuits the evaluation for the vast majority of symbols, completely avoiding the overhead of strncmp() macro expansion. Additionally, since the symbol name could potentially be NULL, we safely guard the character access with an explicit non-null check ((sp)->name &&) to prevent segmentation faults. This patch applies the early reject optimization to: 1. All MODULE_* pseudo-symbol macros. 2. The compare_syms() function for "__insmod" and "_MODULE_START_" checks. This cumulative optimization yields a measurable performance improvement in module symbol sorting and initialization speed. Signed-off-by: Rui Qi <qirui.001(a)bytedance.com> --- symbols.c | 20 ++++++++++---------- 1 file changed, 10 insertions(+), 10 deletions(-) diff --git a/symbols.c b/symbols.c index 736d9d96d606..9b0de45e9a95 100644 --- a/symbols.c +++ b/symbols.c @@ -1314,14 +1314,14 @@ symname_hash_search(struct syment *table[], char *name) * Output for sym -[lL] command. */ -#define MODULE_PSEUDO_SYMBOL(sp) (STRNEQ((sp)->name, "_MODULE_")) +#define MODULE_PSEUDO_SYMBOL(sp) ((sp)->name && (sp)->name[0] == '_' && STRNEQ((sp)->name, "_MODULE_")) -#define MODULE_START(sp) (STRNEQ((sp)->name, "_MODULE_START_")) -#define MODULE_END(sp) (STRNEQ((sp)->name, "_MODULE_END_")) -#define MODULE_INIT_START(sp) (STRNEQ((sp)->name, "_MODULE_INIT_START_")) -#define MODULE_INIT_END(sp) (STRNEQ((sp)->name, "_MODULE_INIT_END_")) -#define MODULE_SECTION_START(sp) (STRNEQ((sp)->name, "_MODULE_SECTION_START")) -#define MODULE_SECTION_END(sp) (STRNEQ((sp)->name, "_MODULE_SECTION_END")) +#define MODULE_START(sp) ((sp)->name && (sp)->name[0] == '_' && STRNEQ((sp)->name, "_MODULE_START_")) +#define MODULE_END(sp) ((sp)->name && (sp)->name[0] == '_' && STRNEQ((sp)->name, "_MODULE_END_")) +#define MODULE_INIT_START(sp) ((sp)->name && (sp)->name[0] == '_' && STRNEQ((sp)->name, "_MODULE_INIT_START_")) +#define MODULE_INIT_END(sp) ((sp)->name && (sp)->name[0] == '_' && STRNEQ((sp)->name, "_MODULE_INIT_END_")) +#define MODULE_SECTION_START(sp) ((sp)->name && (sp)->name[0] == '_' && STRNEQ((sp)->name, "_MODULE_SECTION_START")) +#define MODULE_SECTION_END(sp) ((sp)->name && (sp)->name[0] == '_' && STRNEQ((sp)->name, "_MODULE_SECTION_END")) #define MODULE_MEM_START(sp,t) (STRNEQ((sp)->name, module_tag[t].start)) #define MODULE_MEM_END(sp,t) (STRNEQ((sp)->name, module_tag[t].end)) @@ -3419,12 +3419,12 @@ compare_syms(const void *v1, const void *v2) s2 = (struct syment *)v2; if (s1->value == s2->value) { - if (STRNEQ(s1->name, "__insmod")) + if (s1->name && s1->name[0] == '_' && STRNEQ(s1->name, "__insmod")) return -1; - if (STRNEQ(s2->name, "__insmod")) + if (s2->name && s2->name[0] == '_' && STRNEQ(s2->name, "__insmod")) return 1; if (MODULE_MEM_START(s2, MOD_TEXT) || - STRNEQ(s2->name, "_MODULE_START_")) + s2->name && s2->name[0] == '_' && STRNEQ(s2->name, "_MODULE_START_")) return 1; /* Get pseudo section name. */ if (MODULE_SECTION_START(s1)) -- 2.20.1

2 months

2
2
0 / 0

[PATCH v8 0/3] xarray: add large folio support

by Huang Shijie

The linux kernel supports the large folio for page cache by default. But the current CRASH does not support the large folio. So we may meet the errors when we detected the large folio sometimes, such as in the email: https://www.spinics.net/linux/fedora/redhat-crash-utility/msg11238.html ------------------------------ files: page_to_nid: invalid page: 0 files: page_to_nid: invalid page: 0 files: page_to_nid: invalid page: 0 files: page_to_nid: invalid page: 10 files: page_to_nid: invalid page: 10 files: page_to_nid: invalid page: 10 files: page_to_nid: invalid page: 20 files: page_to_nid: invalid page: 20 files: page_to_nid: invalid page: 20 files: page_to_nid: invalid page: 30 files: page_to_nid: invalid page: 30 files: page_to_nid: invalid page: 30 files: page_to_nid: invalid page: 40 files: page_to_nid: invalid page: 40 files: page_to_nid: invalid page: 40 files: page_to_nid: invalid page: 50 files: page_to_nid: invalid page: 50 files: page_to_nid: invalid page: 50 files: page_to_nid: invalid page: 60 files: page_to_nid: invalid page: 60 ------------------------------ The patches(1,2) are used to add large folio support for CRASH. The patch 3 is newly version of an old patch: it add "files -n" command. v7-->v8: 1.) changed the code in hook ->update_off 2.) Do not use the LINUX version macro. v6-->v7: 1.) Drop the old patch 1. 2.) For patch "xarray: add large folio support": add XARRAY_TYPE_PAGE_CACHE flag for do_xarray() add update_off hook for xarray_ops{} v5-->v6: 1.) Removed the patch for converting ipcs to unix format. 2.) Rewrote the folio_order() 3.) Add hook ->update_count() for do_xarray_info{} 4.) others. v4-->v5: Add a new patch to convert ipcs.c to unix format. Rebased the patch set again. v3-->v4: Fixes the BPF bug by: 1.) patch 1: Add a new type for do_xarray_info 2.) patch 3: Check XARRAY_TYPE_PAGE_CACHE for do_xarray_count()/do_xarray_dump()/do_xarray_dump_cb() v2-->v3: Rewrited the folio_order() in patch 2 to work with different kernel versions. v1-->v2: 1.) Rebase the kernel to later 7.0-rc1(merge window) 2.) Fixed a bug in the patch 3, the latest kernel supports folios whose page order is bigger then 5: "xarray: add large folio support" *** BLURB HERE *** Huang Shijie (3): add folio_order function xarray: add large folio support add "files -n" command for an inode defs.h | 14 +++++++ filesys.c | 111 +++++++++++++++++++++++++++++++++++++++++++++++++----- help.c | 24 +++++++++++- memory.c | 83 +++++++++++++++++++++++++++++++++++++++- symbols.c | 6 +++ tools.c | 15 +++++++- 6 files changed, 238 insertions(+), 15 deletions(-) -- 2.53.0

2 months

2
5
0 / 0

Little Tikkers Watch: Tiny Wrists, Big Style

by Olivia Gormley

I've been kinda searching around for a good-looking but still comfy watch for kids, and these little tikkers watch ideas really caught my attention. The designs seem fun, bright, and they fit those tiny wrists nicely without feeling super bulky or heavy. Has anyone here bought a little tikkers watch for their child lately? I’m mostly wondering about the real quality, how durable it is after a lot of use, and whether kids actually keep it on day after day. Would really like some honest opinions before I buy one, since I don’t want to end up with something that falls apart quickly. Visit Our Website: https://theweartique.com/collections/tikkers-watches-for-kids

2 months

1
0
0 / 0

[PATCH v7 0/3] xarray: add large folio support

by Huang Shijie

The linux kernel supports the large folio for page cache by default. But the current CRASH does not support the large folio. So we may meet the errors when we detected the large folio sometimes, such as in the email: https://www.spinics.net/linux/fedora/redhat-crash-utility/msg11238.html ------------------------------ files: page_to_nid: invalid page: 0 files: page_to_nid: invalid page: 0 files: page_to_nid: invalid page: 0 files: page_to_nid: invalid page: 10 files: page_to_nid: invalid page: 10 files: page_to_nid: invalid page: 10 files: page_to_nid: invalid page: 20 files: page_to_nid: invalid page: 20 files: page_to_nid: invalid page: 20 files: page_to_nid: invalid page: 30 files: page_to_nid: invalid page: 30 files: page_to_nid: invalid page: 30 files: page_to_nid: invalid page: 40 files: page_to_nid: invalid page: 40 files: page_to_nid: invalid page: 40 files: page_to_nid: invalid page: 50 files: page_to_nid: invalid page: 50 files: page_to_nid: invalid page: 50 files: page_to_nid: invalid page: 60 files: page_to_nid: invalid page: 60 ------------------------------ The patches(1,2) are used to add large folio support for CRASH. The patch 3 is newly version of an old patch: it add "files -n" command. v6-->v7: 1.) Drop the old patch 1. 2.) For patch "xarray: add large folio support": add XARRAY_TYPE_PAGE_CACHE flag for do_xarray() add update_off hook for xarray_ops{} v5-->v6: 1.) Removed the patch for converting ipcs to unix format. 2.) Rewrote the folio_order() 3.) Add hook ->update_count() for do_xarray_info{} 4.) others. v4-->v5: Add a new patch to convert ipcs.c to unix format. Rebased the patch set again. v3-->v4: Fixes the BPF bug by: 1.) patch 1: Add a new type for do_xarray_info 2.) patch 3: Check XARRAY_TYPE_PAGE_CACHE for do_xarray_count()/do_xarray_dump()/do_xarray_dump_cb() v2-->v3: Rewrited the folio_order() in patch 2 to work with different kernel versions. v1-->v2: 1.) Rebase the kernel to later 7.0-rc1(merge window) 2.) Fixed a bug in the patch 3, the latest kernel supports folios whose page order is bigger then 5: "xarray: add large folio support" Huang Shijie (3): add folio_order function xarray: add large folio support add "files -n" command for an inode defs.h | 13 +++++++ filesys.c | 108 +++++++++++++++++++++++++++++++++++++++++++++++++----- help.c | 24 +++++++++++- memory.c | 84 +++++++++++++++++++++++++++++++++++++++++- symbols.c | 6 +++ tools.c | 23 +++++++++--- 6 files changed, 240 insertions(+), 18 deletions(-) -- 2.53.0

2 months, 1 week

2
10
0 / 0

[PATCH] Support module memory layout change on linux 6.18

by yangshiguang1011＠163.com

From: yangshiguang <yangshiguang(a)xiaomi.com> Support module memory layout change on linux 6.18 by kernel commit b4760ff ("module: deprecate usage of *_gpl sections in module loader")[1]. Without the patch, crash cannot start a session with an error message like this: crash: invalid structure member offset: module_num_gpl_syms FILE: kernel.c LINE: 3834 FUNCTION: module_init() [1] https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?... Signed-off-by: yangshiguang <yangshiguang(a)xiaomi.com> --- kernel.c | 5 +++-- symbols.c | 18 ++++++++++++++---- 2 files changed, 17 insertions(+), 6 deletions(-) diff --git a/kernel.c b/kernel.c index 8781d6a..786eb73 100644 --- a/kernel.c +++ b/kernel.c @@ -3830,8 +3830,9 @@ module_init(void) nsyms = UINT(modbuf + OFFSET(module_nsyms)); break; case KMOD_V2: - nsyms = UINT(modbuf + OFFSET(module_num_syms)) + - UINT(modbuf + OFFSET(module_num_gpl_syms)); + nsyms = UINT(modbuf + OFFSET(module_num_syms)); + if (VALID_MEMBER(module_num_gpl_syms)) + nsyms += UINT(modbuf + OFFSET(module_num_gpl_syms)); break; } diff --git a/symbols.c b/symbols.c index e6865ca..189fe3e 100644 --- a/symbols.c +++ b/symbols.c @@ -1976,9 +1976,14 @@ store_module_symbols_6_4(ulong total, int mods_installed) "module buffer", FAULT_ON_ERROR); syms = ULONG(modbuf + OFFSET(module_syms)); - gpl_syms = ULONG(modbuf + OFFSET(module_gpl_syms)); nsyms = UINT(modbuf + OFFSET(module_num_syms)); - ngplsyms = UINT(modbuf + OFFSET(module_num_gpl_syms)); + if (VALID_MEMBER(module_gpl_syms) && VALID_MEMBER(module_num_gpl_syms)) { + gpl_syms = ULONG(modbuf + OFFSET(module_gpl_syms)); + ngplsyms = UINT(modbuf + OFFSET(module_num_gpl_syms)); + } else { + gpl_syms = 0; + ngplsyms = 0; + } nksyms = UINT(modbuf + OFFSET(module_num_symtab)); @@ -2336,9 +2341,14 @@ store_module_symbols_v2(ulong total, int mods_installed) "module buffer", FAULT_ON_ERROR); syms = ULONG(modbuf + OFFSET(module_syms)); - gpl_syms = ULONG(modbuf + OFFSET(module_gpl_syms)); nsyms = UINT(modbuf + OFFSET(module_num_syms)); - ngplsyms = UINT(modbuf + OFFSET(module_num_gpl_syms)); + if (VALID_MEMBER(module_gpl_syms) && VALID_MEMBER(module_num_gpl_syms)) { + gpl_syms = ULONG(modbuf + OFFSET(module_gpl_syms)); + ngplsyms = UINT(modbuf + OFFSET(module_num_gpl_syms)); + } else { + gpl_syms = 0; + ngplsyms = 0; + } if (THIS_KERNEL_VERSION >= LINUX(2,6,27)) { nksyms = UINT(modbuf + OFFSET(module_num_symtab)); -- 2.43.0

2 months, 1 week

2
1
0 / 0

[PATCH] RISCV64: introduce riscv64_VTOP() and riscv64_PTOV()

by Austin Kim

The VTOP and PTOV macros are typically implemented using architecture-specific functions to improve maintainability and debugging capabilities. This patch moves the RISCV64-specific VTOP/PTOV logic from inline macros in defs.h to dedicated functions in riscv64.c. This refactoring makes the code easier to understand and provides a better location for adding future debugging or validation checks. No functional changes are introduced by this patch. Signed-off-by: Austin Kim <austindh.kim(a)gmail.com> --- defs.h | 14 ++++---------- riscv64.c | 24 ++++++++++++++++++++++++ 2 files changed, 28 insertions(+), 10 deletions(-) diff --git a/defs.h b/defs.h index a6f4372..54f1725 100644 --- a/defs.h +++ b/defs.h @@ -3803,16 +3803,8 @@ typedef signed int s32; /* * Direct memory mapping */ -#define PTOV(X) \ - (((unsigned long)(X)+(machdep->kvbase)) - machdep->machspec->phys_base) -#define VTOP(X) ({ \ - ulong _X = X; \ - (THIS_KERNEL_VERSION >= LINUX(5,13,0) && \ - (_X) >= machdep->machspec->kernel_link_addr) ? \ - ((unsigned long)(_X)-(machdep->machspec->va_kernel_pa_offset)): \ - (((unsigned long)(_X)-(machdep->kvbase)) + \ - machdep->machspec->phys_base); \ - }) +#define PTOV(X) riscv64_PTOV((ulong)(X)) +#define VTOP(X) riscv64_VTOP((ulong)(X)) #define PAGEBASE(X) (((ulong)(X)) & (ulong)machdep->pagemask) /* @@ -7211,6 +7203,8 @@ void riscv64_display_regs_from_elf_notes(int, FILE *); void riscv64_init(int); void riscv64_dump_machdep_table(ulong); int riscv64_IS_VMALLOC_ADDR(ulong); +ulong riscv64_PTOV(ulong); +ulong riscv64_VTOP(ulong); #define display_idt_table() \ error(FATAL, "-d option is not applicable to RISCV64 architecture\n") diff --git a/riscv64.c b/riscv64.c index eceae70..3aee827 100644 --- a/riscv64.c +++ b/riscv64.c @@ -1606,6 +1606,30 @@ riscv64_uvtop(struct task_context *tc, ulong uvaddr, physaddr_t *paddr, int verb } } +ulong riscv64_PTOV(ulong paddr) +{ + ulong vaddr; + ulong offset = paddr - machdep->machspec->phys_base; + + vaddr = offset + machdep->kvbase; + + return vaddr; +} + +ulong +riscv64_VTOP(ulong addr) +{ + ulong paddr; + + if ( (THIS_KERNEL_VERSION >= LINUX(5,13,0)) && + (addr >= machdep->machspec->kernel_link_addr)) + paddr = (addr - (machdep->machspec->va_kernel_pa_offset)); + else + paddr = (addr - (ulong)machdep->kvbase + machdep->machspec->phys_base); + + return paddr; +} + static int riscv64_kvtop(struct task_context *tc, ulong kvaddr, physaddr_t *paddr, int verbose) { -- 2.34.1

2 months, 1 week

2
5
0 / 0

[PATCH v6 0/4] xarray: add large folio support

by Huang Shijie

The linux kernel supports the large folio for page cache by default. But the current CRASH does not support the large folio. So we may meet the errors when we detected the large folio sometimes, such as in the email: https://www.spinics.net/linux/fedora/redhat-crash-utility/msg11238.html ------------------------------ files: page_to_nid: invalid page: 0 files: page_to_nid: invalid page: 0 files: page_to_nid: invalid page: 0 files: page_to_nid: invalid page: 10 files: page_to_nid: invalid page: 10 files: page_to_nid: invalid page: 10 files: page_to_nid: invalid page: 20 files: page_to_nid: invalid page: 20 files: page_to_nid: invalid page: 20 files: page_to_nid: invalid page: 30 files: page_to_nid: invalid page: 30 files: page_to_nid: invalid page: 30 files: page_to_nid: invalid page: 40 files: page_to_nid: invalid page: 40 files: page_to_nid: invalid page: 40 files: page_to_nid: invalid page: 50 files: page_to_nid: invalid page: 50 files: page_to_nid: invalid page: 50 files: page_to_nid: invalid page: 60 files: page_to_nid: invalid page: 60 ------------------------------ The patches(1,2,3) are used to add large folio support for CRASH. The patch 4 is newly version of an old patch: it add "files -n" command. v5-->v6: 1.) Removed the patch for converting ipcs to unix format. 2.) Rewrote the folio_order() 3.) Add hook ->update_count() for do_xarray_info{} 4.) others. v4-->v5: Add a new patch to convert ipcs.c to unix format. Rebased the patch set again. v3-->v4: Fixes the BPF bug by: 1.) patch 1: Add a new type for do_xarray_info 2.) patch 3: Check XARRAY_TYPE_PAGE_CACHE for do_xarray_count()/do_xarray_dump()/do_xarray_dump_cb() v2-->v3: Rewrited the folio_order() in patch 2 to work with different kernel versions. v1-->v2: 1.) Rebase the kernel to later 7.0-rc1(merge window) 2.) Fixed a bug in the patch 3, the latest kernel supports folios whose page order is bigger then 5: "xarray: add large folio support" *** BLURB HERE *** Huang Shijie (4): xarray: add a new parameter for do_xarray add folio_order function xarray: add large folio support add "files -n" command for an inode bpf.c | 8 ++--- defs.h | 14 ++++++++- dev.c | 4 +-- diskdump.c | 10 ++++-- filesys.c | 90 ++++++++++++++++++++++++++++++++++++++++++++++++------ help.c | 24 ++++++++++++++- ipcs.c | 4 +-- kernel.c | 4 +-- memory.c | 84 ++++++++++++++++++++++++++++++++++++++++++++++++-- symbols.c | 6 ++++ task.c | 4 +-- tools.c | 16 ++++++++-- 12 files changed, 237 insertions(+), 31 deletions(-) -- 2.34.1

2 months, 1 week

2
8
0 / 0

[RESEND PATCH v1 1/1] symbols: Add support for mod symtab with combined GPL and non-GPL symbols

by Alexander Egorenkov

The commit b4760ff2a5e4 ("module: deprecate usage of *_gpl sections in module loader") eliminated separate GPL symbol sections representing GPL only symbols. Therefore, depending on whether the member gpl_syms is present in struct module, decide whether GPL module symbols are separated or not. https://lore.kernel.org/all/20260326-kflagstab-v5-4-fa0796fe88d9@google.com Signed-off-by: Alexander Egorenkov <egorenar(a)linux.ibm.com> --- kernel.c | 13 ++++++++----- symbols.c | 9 +++++++-- 2 files changed, 15 insertions(+), 7 deletions(-) diff --git a/kernel.c b/kernel.c index 8781d6a22414..35e6fd61f077 100644 --- a/kernel.c +++ b/kernel.c @@ -3634,9 +3634,11 @@ module_init(void) case KMOD_V2: MEMBER_OFFSET_INIT(module_num_syms, "module", "num_syms"); MEMBER_OFFSET_INIT(module_list, "module", "list"); - MEMBER_OFFSET_INIT(module_gpl_syms, "module", "gpl_syms"); - MEMBER_OFFSET_INIT(module_num_gpl_syms, "module", - "num_gpl_syms"); + if (MEMBER_EXISTS("module", "gpl_syms")) { + MEMBER_OFFSET_INIT(module_gpl_syms, "module", "gpl_syms"); + MEMBER_OFFSET_INIT(module_num_gpl_syms, "module", + "num_gpl_syms"); + } if (MEMBER_EXISTS("module", "mem")) { /* 6.4 and later */ kt->flags2 |= KMOD_MEMORY; /* MODULE_MEMORY() can be used. */ @@ -3830,8 +3832,9 @@ module_init(void) nsyms = UINT(modbuf + OFFSET(module_nsyms)); break; case KMOD_V2: - nsyms = UINT(modbuf + OFFSET(module_num_syms)) + - UINT(modbuf + OFFSET(module_num_gpl_syms)); + nsyms = UINT(modbuf + OFFSET(module_num_syms)); + if (VALID_MEMBER(module_num_gpl_syms)) + nsyms += UINT(modbuf + OFFSET(module_num_gpl_syms)); break; } diff --git a/symbols.c b/symbols.c index e6865cabef74..7b00e26ab0cf 100644 --- a/symbols.c +++ b/symbols.c @@ -1976,9 +1976,14 @@ store_module_symbols_6_4(ulong total, int mods_installed) "module buffer", FAULT_ON_ERROR); syms = ULONG(modbuf + OFFSET(module_syms)); - gpl_syms = ULONG(modbuf + OFFSET(module_gpl_syms)); nsyms = UINT(modbuf + OFFSET(module_num_syms)); - ngplsyms = UINT(modbuf + OFFSET(module_num_gpl_syms)); + if (VALID_MEMBER(module_gpl_syms)) { + gpl_syms = ULONG(modbuf + OFFSET(module_gpl_syms)); + ngplsyms = UINT(modbuf + OFFSET(module_num_gpl_syms)); + } else { + gpl_syms = 0; + ngplsyms = 0; + } nksyms = UINT(modbuf + OFFSET(module_num_symtab)); -- 2.51.0

2 months, 2 weeks

2
3
0 / 0

[PATCH v2] Fix get_xtime() for kernel 6.13 and higher

by Aureau, Georges (Kernel Tools ERT)

On kernels 6.13 and higher, DATE is not displayed properly: crash> sys | grep -e DATE -e REL DATE: Wed Dec 31 18:00:00 CST 1969 RELEASE: 7.0.0-14-generic The function get_xtime() was enhanced to support: - kernel 6.13 ("shadow_timekeeper" moved to "struct tk_data tk_core") - kernel 6.17 ("tk_core" is now "timekeeper_data[0]). With this get_xtime() fix: crash> sys | grep DATE DATE: Mon Apr 27 09:19:50 CDT 2026 Signed-off-by: Georges Aureau <georges.aureau(a)hpe.com> -- defs.h | 1 + kernel.c | 15 ++++++++++++++- 2 files changed, 15 insertions(+), 1 deletion(-) diff --git a/defs.h b/defs.h index a6f4372..c07b9a2 100644 --- a/defs.h +++ b/defs.h @@ -2290,6 +2290,7 @@ struct offset_table { /* stash of commonly-used offsets */ long bpf_ringbuf_nr_pages; long hrtimer_clock_base_index; long klp_patch_list; + long tk_data_timekeeper; }; struct size_table { /* stash of commonly-used sizes */ diff --git a/kernel.c b/kernel.c index 8781d6a..0dafe74 100644 --- a/kernel.c +++ b/kernel.c @@ -258,6 +258,7 @@ kernel_init() MEMBER_OFFSET_INIT(timekeeper_xtime, "timekeeper", "xtime"); MEMBER_OFFSET_INIT(timekeeper_xtime_sec, "timekeeper", "xtime_sec"); + MEMBER_OFFSET_INIT(tk_data_timekeeper, "tk_data", "timekeeper"); get_xtime(&kt->date); if (CRASHDEBUG(1)) fprintf(fp, "xtime timespec.tv_sec: %lx: %s\n", @@ -11134,7 +11135,19 @@ get_xtime(struct timespec *date) struct syment *sp; uint64_t xtime_sec; - if (VALID_MEMBER(timekeeper_xtime) && + if (VALID_MEMBER(tk_data_timekeeper) && + VALID_MEMBER(timekeeper_xtime_sec)) { + long offset = OFFSET(tk_data_timekeeper) + + OFFSET(timekeeper_xtime_sec); + if ((sp = kernel_symbol_search("timekeeper_data")) || + (sp = kernel_symbol_search("tk_core"))) { + readmem(sp->value + offset, KVADDR, + &xtime_sec, sizeof(uint64_t), + "tk_data timekeeper xtime_sec", RETURN_ON_ERROR); + date->tv_sec = (__time_t)xtime_sec; + } + } + else if (VALID_MEMBER(timekeeper_xtime) && (sp = kernel_symbol_search("timekeeper"))) { readmem(sp->value + OFFSET(timekeeper_xtime), KVADDR, date, sizeof(struct timespec),

2 months, 2 weeks

2
2
0 / 0

[PATCH 0/2] RISCV64: optimize symbol handling for large kernel symbol tables

by Rui Qi

This patch series optimizes symbol handling in crash utility for RISCV64 architecture, particularly beneficial when debugging kernels with large symbol tables (900K+ symbols). Background: On a 32-core RISCV64 physical machine running Linux 6.12.13, the kernel symbol table contains over 900,000 entries (wc -l /proc/kallsyms). When using crash utility to analyze vmcore or live system, significant time is spent in symbol-related operations, specifically: 1. Symbol name hash lookups - the original simple hash function causes high collision rates with large symbol tables 2. Processing of linker-generated mapping symbols (.L*, L0*, $*) which are unnecessary for debugging but consume processing time Patch 1: Optimize symname_hash with larger table and FNV-1a hash - Increases SYMNAME_HASH from 512 to 16384 (32x) to reduce collisions - Replaces simple hash with FNV-1a algorithm for better distribution - Removes strlen() call by computing hash in single pass This reduces average hash bucket chain length significantly, improving symbol lookup performance at the cost of ~248KB additional memory. Patch 2: Add mapping symbol filter in riscv64_verify_symbol - Filters out linker mapping symbols (.L*, L0*, $*) that should not be in the symbol list - Also optimizes riscv64_verify_symbol() by consolidating name validity checks Together these patches improve crash utility startup and symbol lookup performance on RISCV64 systems with large kernel symbol tables. Rui Qi (2): symbols: optimize symname_hash with larger table and FNV-1a hash RISCV64: add mapping symbol filter in riscv64_verify_symbol defs.h | 2 +- riscv64.c | 14 +++++++++++--- symbols.c | 20 ++++++++++++++------ 3 files changed, 27 insertions(+), 9 deletions(-) -- 2.43.0

2 months, 2 weeks

2
4
0 / 0

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

Crash-utility May 2026