Re: [PATCH 2/2] Add LoongArch64 basic support

On 2024-07-20 03:48, Camm Maguire wrote:

> I cannot reproduce this on cfarm400 -- can you please post the output of
> these tests earlier in the log?
>
> Should be no test failures virtually everywhere, except a few platforms
> which have trouble trapping floating point exceptions (exp and expt test
> failures).
It is glibc::localtime causing failed.

Test DECODE-UNIVERSAL-TIME.1 failed
Form: (DECODE-UNIVERSAL-TIME 0 0)
Expected values: 0
                  0
                  0
                  1
                  1
                  1900
                  0
                  NIL
                  0
Actual values: 43
                5
                0
                1
                1
                1900
                0
                NIL
                0.

I test localtime by C,

#include <stdio.h>
#include <time.h>

void printTM(struct tm *timeptr) {
     printf("tm_sec: %d\n", timeptr->tm_sec);
     printf("tm_min: %d\n", timeptr->tm_min);
     printf("tm_hour: %d\n", timeptr->tm_hour);
     printf("tm_mday: %d\n", timeptr->tm_mday);
     printf("tm_mon: %d\n", timeptr->tm_mon);
     printf("tm_year: %d\n", timeptr->tm_year);
     printf("tm_wday: %d\n", timeptr->tm_wday);
     printf("tm_yday: %d\n", timeptr->tm_yday);
     printf("tm_isdst: %d\n", timeptr->tm_isdst);
}

int main() {
     struct tm * timeinfo;
     time_t raw = 0;
     raw -= (time_t)60 * 60 * 24 * (70 * 365 + 17);
     raw -= (time_t)8 * 60 * 60;  // Asia/Shanghai
     timeinfo = localtime(&raw);
     printTM(timeinfo);
     return 0;
}

The result is,
tm_sec: 43
tm_min: 5
tm_hour: 0
tm_mday: 1
tm_mon: 0
tm_year: 0
tm_wday: 1
tm_yday: 0
tm_isdst: 0


Thank you again. The gcl27 has been successfully built by Debian 
autobuilder.

Jinyang