[PATCH] newlib: libc: Improved the readability of strcspn with minor optimization

[Xiao Zeng]

2023-12-16 12:45  Xiao Zeng <zengxiao@eswincomputing.com> wrote:
>
 
>2023-12-15 18:28  Torbjorn SVENSSON <torbjorn.svensson@foss.st.com> wrote:
>>
>Hi Bob, I worked under the risc-v architecture and obtained the newlib library through 
Sorry, Torbjorn, I accidentally wrote your name wrong. I apologize for my carelessness.

>cross compilation. This is my first attempt to contribute code to newlib. 
>
>>Hello Xiao,
>>
>>On 2023-12-15 09:31, Xiao Zeng wrote:
>>> Signed-off-by: Xiao Zeng <zengxiao@eswincomputing.com>
>>> ---
>>>   newlib/libc/string/strcspn.c | 6 ++----
>>>   1 file changed, 2 insertions(+), 4 deletions(-)
>>>
>>> diff --git a/newlib/libc/string/strcspn.c b/newlib/libc/string/strcspn.c
>>> index abaa93ad6..8ac0bf10c 100644
>>> --- a/newlib/libc/string/strcspn.c
>>> +++ b/newlib/libc/string/strcspn.c
>>> @@ -37,12 +37,10 @@ strcspn (const char *s1,
>>>         for (c = s2; *c; c++)
>>>   {
>>>     if (*s1 == *c)
>>> -	    break;
>>> +	    goto end;
>>>   }
>>> -      if (*c)
>>> -	break;
>>>         s1++;
>>>       }
>>> -
>>> +end:
>>>     return s1 - s;
>>>   }
>>
>>Just looking at this small snippet of code, I would say that the
>>previous code and your suggestion won't do the same thing.
>>
>>Do you have unit tests that confirm that the behavior is identical with
>>the current implementation and your suggested change?
>1 I must admit that I did not conduct a complete newlib regression test.
>Anyway, I should apologize for this. After completing cross compilation, I tried:
>-----------------
>make check
>-----------------
>
>Received the following error message:
>--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
>Target is riscv64-unknown-elf
>Host   is x86_64-pc-linux-gnu
>...
>The error code is TCL LOOKUP COMMAND newlib_target_compile
>The info on the error is:
>invalid command name "newlib_target_compile"
>    while executing
>"::tcl_unknown newlib_target_compile /home/user/Downloads/tools/tools.git/newlib.git/newlib/testsuite/newlib.iconv/iconvnm.c /home/user/Downloads/tools..."
>    ("uplevel" body line 1)
>    invoked from within
>"uplevel 1 ::tcl_unknown $args"
>--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
>
>I know the reason for the error: Target and Host do not match, which
>is a common issue in cross compilation. I have looked up some
>information, but I have not found a solution.
>
>2 On the basis of the above, I searched for test cases of the strcspn
>function in newlib:
>-------------------------------------------------------------------------------------------------
>newlib/libm/test/string.c:311:  check(strcspn("abcba", "qx") == 5); /* Whole string. */
>newlib/libm/test/string.c:312:  check(strcspn("abcba", "cx") == 2); /* Partial. */
>newlib/libm/test/string.c:313:  check(strcspn("abc", "abc") == 0);	/* None. */
>newlib/libm/test/string.c:314:  check(strcspn("", "ab") == 0); /* Null string. */
>newlib/libm/test/string.c:315:  check(strcspn("abc", "") == 3); /* Null search list. */
>-------------------------------------------------------------------------------------------------
>
>I simply believe that these 5 test cases are all the test cases related to
>strcspn in newlib. After local testing, all these test cases can pass in my patch.
>>
>>When I run your suggestion, I get return value 0, but with the current
>>implementation it's 3 for this call: strspn("129th", "1234567890").
>3 It is worth noting that strcspn and strspn only differ by one letter 'c'. I also
>often confuse them -:), maybe you are the same.
>
>>
>>Kind regards,
>>Torbjörn
> 
>4 Perhaps someone could point out how to test newlib under the risc-v
>architecture, and I would greatly appreciate it. Of course, there are also
>great methods for testing newlib under other architectures.
>
>5 By the way, it is not possible to compile the latest newlib on host (x86_64-pc-linux-gnu)
>(compilation quickly ended as if nothing had happened), even if ../configure
>comes with the parameter -- with-newlib.
>
>
>Thanks
>Xiao Zeng
>
 
Thanks
Xiao Zeng