Re: [PATCH 1/2] perf arm64: Handle __NR3264_ prefixed syscall number

[Tiezhu Yang <yangtiezhu@loongson.cn>]

On 05/24/2023 02:43 PM, Alexander Kapshuk wrote:
> On Wed, May 24, 2023 at 6:19 AM Tiezhu Yang <yangtiezhu@loongson.cn> wrote:
>>
>>
>>
>> On 05/23/2023 08:31 PM, Alexander Kapshuk wrote:
>>> On Tue, May 23, 2023 at 1:22 PM Tiezhu Yang <yangtiezhu@loongson.cn> wrote:
>>>>
>>>> After commit 9854e7ad35fe ("perf arm64: Simplify mksyscalltbl"),
>>>> in the generated syscall table file syscalls.c, there exist some
>>>> __NR3264_ prefixed syscall numbers such as [__NR3264_ftruncate],
>>>> it looks like not so good, just do some small filter operations
>>>> to handle __NR3264_ prefixed syscall number as a digital number.
>>>>
>>>> Without this patch:
>>>>
>>>>   [__NR3264_ftruncate] = "ftruncate",
>>>>
>>>> With this patch:
>>>>
>>>>   [46] = "ftruncate",
>>>>
>>>> Signed-off-by: Tiezhu Yang <yangtiezhu@loongson.cn>
>>>> ---
>>>>  tools/perf/arch/arm64/entry/syscalls/mksyscalltbl | 7 ++++---
>>>>  1 file changed, 4 insertions(+), 3 deletions(-)
>>>>
>>>> diff --git a/tools/perf/arch/arm64/entry/syscalls/mksyscalltbl b/tools/perf/arch/arm64/entry/syscalls/mksyscalltbl
>>>> index 22cdf91..59ab7939 100755
>>>> --- a/tools/perf/arch/arm64/entry/syscalls/mksyscalltbl
>>>> +++ b/tools/perf/arch/arm64/entry/syscalls/mksyscalltbl
>>>> @@ -39,7 +39,8 @@ create_table()
>>>>         echo "};"
>>>>  }
>>>>
>>>> -$gcc -E -dM -x c -I $incpath/include/uapi $input \
>>>> -       |sed -ne 's/^#define __NR_//p' \
>>>> -       |sort -t' ' -k2 -n             \
>>>> +$gcc -E -dM -x c -I $incpath/include/uapi $input               \
>>>> +       |awk '{if ($2~"__NR" && $3 !~"__NR3264_") {print}}'     \
>>>> +       |sed -ne 's/^#define __NR_//p;s/^#define __NR3264_//p'  \
>>>> +       |sort -t' ' -k2 -n                                      \
>>>>         |create_table
>>>> --
>>>> 2.1.0
>>>>
>>>
>>> As an aside, the awk + sed + sort parts of the command line may be
>>> reduced to the following awk script, if desired:
>>> awk '$2 ~ "__NR" && $3 !~ "__NR3264_" {
>>>         sub("^#define __NR_", "")
>>>         sub("^#define __NR3264_", "")
>>>         print | "sort -k2 -n"
>>> }'
>>>
>>
>> Hi Alexander,
>>
>> Thanks, it seems more simple and works well as expected.
>> Let us wait for more review comments before respin.
>>
>> If no any objections, I will send v2 with the following
>> changes based on the current patch in the next week.
>>
>> -$gcc -E -dM -x c -I $incpath/include/uapi $input               \
>> -       |awk '{if ($2~"__NR" && $3 !~"__NR3264_") {print}}'     \
>> -       |sed -ne 's/^#define __NR_//p;s/^#define __NR3264_//p'  \
>> -       |sort -t' ' -k2 -n                                      \
>> +$gcc -E -dM -x c -I $incpath/include/uapi $input \
>> +       |awk '$2 ~ "__NR" && $3 !~ "__NR3264_" {
>> +               sub("^#define __NR_", "")
>> +               sub("^#define __NR3264_", "")
>> +               print | "sort -k2 -n"}' \
>>          |create_table
>>
>> Thanks,
>> Tiezhu
>>
>
> Hi Tiezhu,
>
> Thanks for your prompt feedback.
> It was merely a suggestion entirely subject to your discretion.
>
> If no other patterns are anticipated to be processed by the sub

Yes, there are only 2 patterns such as "__NR_" and "__NR3264_",
I confirmed that in include/uapi/asm-generic/unistd.h.

> routines, they may be combined into a single sub routine like so:
> awk '$2 ~ "__NR" && $3 !~ "__NR3264_" {
>         sub("^#define __NR(3264)?_", "")
>         print | "sort -k2 -n"
> }'

Thanks again, I tested the above code, it also works well and
looks better, I will modify the code as you suggested in v2.

Thanks,
Tiezhu