I will take a look.

On 20/06/25 07:16, Maxim Kuvyrkov wrote:

Hi Sam,

Thanks for looking into this.

Hi Adhemerval,

Would you please look into how glibc should handle this?

Thanks!

-- Maxim Kuvyrkov https://www.linaro.org

...

On Jun 19, 2025, at 04:24, Sam James sam@gentoo.org wrote:

Could you file a glibc bug for this please? It needs some analysis to see if it's an FP or not.

There's actually a suppression for -Os already there:

                       if (elem < table_size)
                         {
                           /* Compare the byte sequence but only if
                              this is not part of a range.  */

                           /* The compiler might warn that idx may be
                              used uninitialized, however it will be
                              reached iff elem < table_size which means
                              that it was properly set in the loop
                              above.   */
                           DIAG_PUSH_NEEDS_COMMENT;
                           DIAG_IGNORE_Os_NEEDS_COMMENT (8, "-Wmaybe-uninitialized");
                           if (! is_range

# if WIDE_CHAR_VERSION && __wmemcmp (n, &wextra[1], c1) == 0 # else && memcmp (n, &extra[idx + 1], c1) == 0 # endif ) { n += c1 - 1; goto matched; } DIAG_POP_NEEDS_COMMENT; _______________________________________________ linaro-toolchain mailing list -- linaro-toolchain@lists.linaro.org To unsubscribe send an email to linaro-toolchain-leave@lists.linaro.org