10:48 p.m.
Hi,
The crash utility could not get module percpu symbols.
Since percpu symbols are not included in kallsyms nor module_core area,
they were not found out in module_init().
However, symbols are existing in module object file
.data..percpu (legacy kernel revision .data.percpu) section,
load_module_symbols() can find and resolve them.
(percpu symbols exist in mod_load_symtable, not in mod_ext_symtable)
'p' command is also updated to be made redundant legacy per_cpu__ prefix.
Note:
I have not tested these paches enough over any arch or kernel revisions,
also any distros.
Just have been tested x86 (64bit) with linux-2.6.10 and linux-2.6.35.
This is my first post to crash utility.
I don't understand well the manner of this mailing list.
(I'm sorry if there's anything wrong)
Thanks,
Toshi.
-----------------------------------------------------------------------------
CHANGES: module percpu symbols are resolved while module object loading.
cmd_p() can read percpu syombols without prefix in legacy environs.
SRPM: crash-5.0.9-0.src.rpm
TEST#1: Result on linux-2.6.10 x86(64) SMP(4 CPUs) * with test module
[before]
crash> sym -m test | grep abc
crash> p abc
p: gdb request failed: p abc
[after]
crash> sym -m test | grep abc
ffffffff80871928 (d) per_cpu__abc
crash> sym abc
symbol not found: abc
possible alternatives:
ffffffff80871928 (d) per_cpu__abc
crash> p abc
PER-CPU DATA TYPE:
int per_cpu__abc;
PER-CPU ADDRESSES:
[0]: 10008af0888
[1]: 10008b00888
[2]: 10008b10888
[3]: 10008b20888
TEST#2: Result on linux-2.6.35 x86(64) SMP (8 CPUs) * with KVM modules
crash> mod | grep kvm
ffffffffa0035c40 kvm_intel 39368
/lib/modules/2.6.35/kernel/arch/x86/kvm/kvm-intel.ko
ffffffffa008dba0 kvm 175160 /lib/modules/2.6.35/kernel/arch/x86/kvm/kvm.ko
[before]
crash> sym -m kvm_intel | grep vcpus_on_cpu
crash> p vcpus_on_cpu
p: gdb request failed: p vcpus_on_cpu
[after]
crash> sym -m kvm_intel | grep vcpus_on_cpu
12a40 (d) vcpus_on_cpu
crash> p vcpus_on_cpu
PER-CPU DATA TYPE:
struct list_head vcpus_on_cpu;
PER-CPU ADDRESSES:
[0]: ffff880001812a40
[1]: ffff880001852a40
[2]: ffff880001892a40
[3]: ffff8800018d2a40
[4]: ffff880001912a40
[5]: ffff880001952a40
[6]: ffff880001992a40
[7]: ffff8800019d2a40
Toshikazu Nakayama (4):
Add percpu member.
module's percpu basic procedure.
expand percpu area.
update cmd_p().
defs.h | 2 ++
kernel.c | 1 +
symbols.c | 50 +++++++++++++++++++++++++++++++++++++++++++++-----
3 files changed, 48 insertions(+), 5 deletions(-)
--
1.7.3.2.161.g3089c
10:57 p.m.
New subject: [PATCH 1/4] Add percpu member.
Add module percpu in load_module and offset_table struct.
This is used for KMOD_V2.
Signed-off-by: Toshikazu Nakayama <nakayama.ts(a)ncos.nec.co.jp>
---
defs.h | 2 ++
1 files changed, 2 insertions(+), 0 deletions(-)
diff --git a/defs.h b/defs.h
index dcece4a..3b279ae 100755
--- a/defs.h
+++ b/defs.h
@@ -1090,6 +1090,7 @@ struct offset_table { /* stash of commonly-used
offsets */
long module_num_symtab;
long module_symtab;
long module_strtab;
+ long module_percpu;
long module_kallsyms_start;
long kallsyms_header_sections;
@@ -2063,6 +2064,7 @@ struct load_module {
ulong mod_init_size;
struct syment *mod_init_symtable;
struct syment *mod_init_symend;
+ ulong mod_percpu;
};
#define IN_MODULE(A,L) \
--
1.7.3.2.161.g3089c
10:58 p.m.
New subject: [PATCH 2/4] module's percpu basic procedure.
- Initialize new member at module_init()
- Make static SEC_FOUND to .data..percpu and .data.percpu sections
at store_section_data()
These can not read from /dev/mem or core file.
(not include in kallsyms nor module_core.)
- Install syms into st->mod_load_namespace at store_load_module_symbols()
module's percpu section include bfd of module object.
The syment.value is sum of bfd value and module.percpu.
Signed-off-by: Toshikazu Nakayama <nakayama.ts(a)ncos.nec.co.jp>
---
kernel.c | 1 +
symbols.c | 12 +++++++++++-
2 files changed, 12 insertions(+), 1 deletions(-)
diff --git a/kernel.c b/kernel.c
index 9ba0dc4..3439c5a 100755
--- a/kernel.c
+++ b/kernel.c
@@ -2694,6 +2694,7 @@ module_init(void)
MEMBER_OFFSET_INIT(module_init_size, "module", "init_size");
MEMBER_OFFSET_INIT(module_init_text_size, "module",
"init_text_size");
+ MEMBER_OFFSET_INIT(module_percpu, "module", "percpu");
/*
* Make sure to pick the kernel "modules" list_head symbol,
diff --git a/symbols.c b/symbols.c
index ec9c893..0a1d170 100755
--- a/symbols.c
+++ b/symbols.c
@@ -1349,6 +1349,7 @@ store_module_symbols_v2(ulong total, int mods_installed)
lm->mod_ext_symcnt = mcnt;
lm->mod_init_module_ptr = ULONG(modbuf +
OFFSET(module_module_init));
+ lm->mod_percpu = ULONG(modbuf + OFFSET(module_percpu));
if (THIS_KERNEL_VERSION >= LINUX(2,6,27)) {
lm->mod_etext_guess = lm->mod_base +
UINT(modbuf + OFFSET(module_core_text_size));
@@ -8346,6 +8347,10 @@ store_section_data(struct load_module *lm, bfd *bfd, asection
*section)
lm->mod_section_data[i].section = section;
lm->mod_section_data[i].priority = prio;
lm->mod_section_data[i].flags = section->flags & ~SEC_FOUND;
+ /* percpu section isn't included in kallsyms nor module_core area. */
+ if (lm->mod_percpu &&
+ (STREQ(name,".data.percpu") || STREQ(name, ".data..percpu")))
+ lm->mod_section_data[i].flags |= SEC_FOUND;
lm->mod_section_data[i].size = bfd_section_size(bfd, section);
lm->mod_section_data[i].offset = 0;
if (strlen(name) < MAX_MOD_SEC_NAME)
@@ -9314,7 +9319,12 @@ store_load_module_symbols(bfd *bfd, int dynamic, void *minisyms,
else if ((spx = kallsyms_module_symbol(lm, &syminfo))) {
syminfo.value = spx->value;
found = TRUE;
- } else {
+ } else if (lm->mod_percpu &&
+ (STREQ(secname, ".data.percpu") ||
+ STREQ(secname, ".data..percpu"))) {
+ syminfo.value += lm->mod_percpu;
+ found = TRUE;
+ } else {
syminfo.value +=
lm->mod_section_data[i].offset + lm->mod_base;
found = TRUE;
}
--
1.7.3.2.161.g3089c
10:58 p.m.
New subject: [PATCH 3/4] expand percpu area.
Current st->__per_cpu_end points at kernel's __per_cpu_end.
Since insmod require new percpu area for target module,
expand st->__per_cpu_end to include module's percpu area,
Replace from symbol_value("sym") to st->sym at display_per_cpu_info()
because they have already been initialized.
Signed-off-by: Toshikazu Nakayama <nakayama.ts(a)ncos.nec.co.jp>
---
symbols.c | 33 ++++++++++++++++++++++++++++++---
1 files changed, 30 insertions(+), 3 deletions(-)
diff --git a/symbols.c b/symbols.c
index 0a1d170..4b4a0e3 100755
--- a/symbols.c
+++ b/symbols.c
@@ -1271,8 +1271,12 @@ store_module_symbols_v2(ulong total, int mods_installed)
char buf3[BUFSIZE];
char buf4[BUFSIZE];
char *strbuf, *modbuf, *modsymbuf;
- struct syment *sp;
+ struct syment *sp, *sp2;
ulong first, last;
+ uint pcpu_num_used;
+ ulong pcpu_size;
+ uint block_size;
+ int *pcpu_vals;
st->mods_installed = mods_installed;
@@ -1621,6 +1625,29 @@ store_module_symbols_v2(ulong total, int mods_installed)
}
}
+ if (kt->flags & KALLSYMS_V2) {
+ /* expand percpu range for module. */
+ if (((sp = symbol_search("pcpu_num_used")) && sp->value)
&&
+ ((sp2 = symbol_search("pcpu_size")))) {
+ readmem(sp->value, KVADDR, &pcpu_num_used,
+ sizeof(pcpu_num_used), "percpu used",
+ FAULT_ON_ERROR);
+ readmem(sp2->value, KVADDR, &pcpu_size,
+ sizeof(pcpu_size), "percpu size ptr",
+ FAULT_ON_ERROR);
+ pcpu_vals = (int *)malloc(sizeof(int) * pcpu_num_used);
+ readmem(pcpu_size, KVADDR, pcpu_vals,
+ sizeof(int) * pcpu_num_used, "percpu size vals",
+ FAULT_ON_ERROR);
+ block_size = 0;
+ for (i = 0; i < pcpu_num_used; i++)
+ block_size += abs(pcpu_vals[i]);
+ free(pcpu_vals);
+ st->__per_cpu_end += block_size;
+ } else if ((sp = symbol_search("pcpu_reserved_chunk_limit")))
+ st->__per_cpu_end = sp->value;
+ }
+
st->flags |= MODULE_SYMS;
if (symbol_query("__insmod_", NULL, NULL))
@@ -5699,8 +5726,8 @@ display_per_cpu_info(struct syment *sp)
char buf[BUFSIZE];
if (((kt->flags & (SMP|PER_CPU_OFF)) != (SMP|PER_CPU_OFF)) ||
- (sp->value < symbol_value("__per_cpu_start")) ||
- (sp->value >= symbol_value("__per_cpu_end")) ||
+ (sp->value < st->__per_cpu_start) ||
+ (sp->value >= st->__per_cpu_end) ||
!((sp->type == 'd') || (sp->type == 'D') || (sp->type ==
'V')))
return FALSE;
--
1.7.3.2.161.g3089c
10:59 p.m.
New subject: [PATCH 4/4] update cmd_p().
crash> p hoge => failed
p: gdb request failed: p hoge
crash> p per_cpu__hoge => success
PER-CPU DATA TYPE:
long unsigned int per_cpu__hoge;
PER-CPU ADDRESSES:
[0]: 10008af0890
:
:
When symbol_search() was failed, did not try per_cpu_symbol_search().
Try it for legacy percpu symbol (per_cpu__ prefix can disappear).
Signed-off-by: Toshikazu Nakayama <nakayama.ts(a)ncos.nec.co.jp>
---
symbols.c | 5 ++++-
1 files changed, 4 insertions(+), 1 deletions(-)
diff --git a/symbols.c b/symbols.c
index 4b4a0e3..b883e83 100755
--- a/symbols.c
+++ b/symbols.c
@@ -5666,7 +5666,10 @@ cmd_p(void)
leader = strlen(buf2);
if (module_symbol(sp->value, NULL, NULL, NULL, *gdb_output_radix))
do_load_module_filter = TRUE;
- } else if (st->flags & LOAD_MODULE_SYMS)
+ } else if ((percpu_sp = per_cpu_symbol_search(args[optind])) &&
+ display_per_cpu_info(percpu_sp))
+ return;
+ else if (st->flags & LOAD_MODULE_SYMS)
do_load_module_filter = TRUE;
if (leader || do_load_module_filter)
--
1.7.3.2.161.g3089c
4989
days inactive
4989
days old
devel@lists.crash-utility.osci.io
4 comments
1 participants
participants (1)
-
Toshikazu Nakayama