| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <ZsiL6aW-nSxCvYBP@x1> Date: Fri, 23 Aug 2024 10:17:29 -0300 From: Arnaldo Carvalho de Melo <acme@...nel.org> To: Howard Chu <howardchu95@...il.com> Cc: adrian.hunter@...el.com, irogers@...gle.com, jolsa@...nel.org, kan.liang@...ux.intel.com, namhyung@...nel.org, linux-perf-users@...r.kernel.org, linux-kernel@...r.kernel.org Subject: Re: [PATCH v2 04/10] perf trace: Add some string arguments' name in syscall_arg_fmt__init_array() On Fri, Aug 23, 2024 at 12:37:01PM +0800, Howard Chu wrote: > On Fri, Aug 23, 2024 at 6:14 AM Arnaldo Carvalho de Melo > <acme@...nel.org> wrote: > > > > On Thu, Aug 15, 2024 at 09:36:20AM +0800, Howard Chu wrote: > > > Add them so that we can augment more strings (which is a file path) > > > > > > Signed-off-by: Howard Chu <howardchu95@...il.com> > > > --- > > > tools/perf/builtin-trace.c | 8 +++++++- > > > 1 file changed, 7 insertions(+), 1 deletion(-) > > > > > > diff --git a/tools/perf/builtin-trace.c b/tools/perf/builtin-trace.c > > > index e7e8c89d9538..84c7398312d8 100644 > > > --- a/tools/perf/builtin-trace.c > > > +++ b/tools/perf/builtin-trace.c > > > @@ -1918,7 +1918,13 @@ syscall_arg_fmt__init_array(struct syscall_arg_fmt *arg, struct tep_format_field > > > > > > if (strcmp(field->type, "const char *") == 0 && > > > ((len >= 4 && strcmp(field->name + len - 4, "name") == 0) || > > > - strstr(field->name, "path") != NULL)) > > > + strstr(field->name, "path") || > > > + strstr(field->name, "file") || > > > + strstr(field->name, "root") || > > > + strstr(field->name, "key") || > > > + strstr(field->name, "special") || > > > + strstr(field->name, "type") || > > > + strstr(field->name, "description"))) > > > arg->scnprintf = SCA_FILENAME; > > > else if ((field->flags & TEP_FIELD_IS_POINTER) || strstr(field->name, "addr")) > > > arg->scnprintf = SCA_PTR; > > > > Humm? > > > > root@...ber:~# for field_name in file root key special type description ; do grep "field:.* $field_name\>" /sys/kernel/tracing/events/syscalls/sys_enter_*/format ; done > > > > /sys/kernel/tracing/events/syscalls/sys_enter_msgget/format: field:key_t key; offset:16; size:8; signed:0; > > /sys/kernel/tracing/events/syscalls/sys_enter_semget/format: field:key_t key; offset:16; size:8; signed:0; > > /sys/kernel/tracing/events/syscalls/sys_enter_shmget/format: field:key_t key; offset:16; size:8; signed:0; > > /sys/kernel/tracing/events/syscalls/sys_enter_quotactl/format: field:const char * special; offset:24; size:8; signed:0; > > /sys/kernel/tracing/events/syscalls/sys_enter_kcmp/format: field:int type; offset:32; size:8; signed:0; > > /sys/kernel/tracing/events/syscalls/sys_enter_mount/format: field:char * type; offset:32; size:8; signed:0; > > /sys/kernel/tracing/events/syscalls/sys_enter_socket/format: field:int type; offset:24; size:8; signed:0; > > /sys/kernel/tracing/events/syscalls/sys_enter_socketpair/format: field:int type; offset:24; size:8; signed:0; > > /sys/kernel/tracing/events/syscalls/sys_enter_syslog/format: field:int type; offset:16; size:8; signed:0; > > root@...ber:~# > > > > Skipping this one. Please ellaborate, what am I missing? > > Hello, just some minor changes on your command, if I do: > ``` > perf $ for field_name in file root key special type description ; do > grep "field:.*char \* .*$field_name\>" > /sys/kernel/tracing/events/syscalls/sys_enter_*/format ; done > /sys/kernel/tracing/events/syscalls/sys_enter_swapoff/format: > field:const char * specialfile; offset:16; size:8; signed:0; > /sys/kernel/tracing/events/syscalls/sys_enter_swapon/format: > field:const char * specialfile; offset:16; size:8; signed:0; > /sys/kernel/tracing/events/syscalls/sys_enter_pivot_root/format: > field:const char * new_root; offset:16; size:8; signed:0; > /sys/kernel/tracing/events/syscalls/sys_enter_fsconfig/format: > field:const char * _key; offset:32; size:8; signed:0; > /sys/kernel/tracing/events/syscalls/sys_enter_quotactl/format: > field:const char * special; offset:24; size:8; signed:0; > /sys/kernel/tracing/events/syscalls/sys_enter_add_key/format: > field:const char * _type; offset:16; size:8; signed:0; > /sys/kernel/tracing/events/syscalls/sys_enter_mount/format: > field:char * type; offset:32; size:8; signed:0; > /sys/kernel/tracing/events/syscalls/sys_enter_request_key/format:field:const > char * _type; offset:16; size:8; signed:0; > /sys/kernel/tracing/events/syscalls/sys_enter_add_key/format: > field:const char * _description; offset:24; size:8; signed:0; > /sys/kernel/tracing/events/syscalls/sys_enter_request_key/format:field:const > char * _description; offset:24; size:8; signed:0; > ``` > > They pop up. > > Because it's strstr(), not strcmp(). Do you think we should use Sure, that was my mistake, I was looking for those exact words, and its being used as substrings. > "strstr(field->name, "description") ||" or "strstr(field->name, > "_description") ||"? Please let me know. But I think using "key" as a a substring is way too generic to think that a syscall arg with that in its name is necessarily a string. We better look at the tracepoint _type_ instead since we have it in the first place :-) - Arnaldo
Powered by blists - more mailing lists