On Fri, 28 Nov 2025, LIU Hao wrote:
This uses the same pattern with the other non-printf128 path. Here digits
are written backwards i.e. from the least significant end to the most
significant end, so they are counted as `p` is increased i.e. in humans'
order from right to left.
Signed-off-by: LIU Hao <[email protected]>
---
mingw-w64-crt/stdio/mingw_pformat.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/mingw-w64-crt/stdio/mingw_pformat.c
b/mingw-w64-crt/stdio/mingw_pformat.c
index fc7429ba4..25ceffb50 100644
--- a/mingw-w64-crt/stdio/mingw_pformat.c
+++ b/mingw-w64-crt/stdio/mingw_pformat.c
@@ -779,8 +779,8 @@ void __pformat_int( __pformat_intarg_t value, __pformat_t
*stream )
memset(p,0,bufflen);
for(int32_t i = strlen(tmp_buff) - 1; i >= 0; i--){
- if ( i && (stream->flags & PFORMAT_GROUPED) != 0 && stream->thousands_chr
!= 0
- && (i % 4) == 3)
+ if (p != buf && (stream->flags & PFORMAT_GROUPED) != 0 &&
stream->thousands_chr != 0
+ && ((p - buf) % 4) == 3)
{
*p++ = ',';
}
--
Same as for the other patch; I can't really say by just looking at the
patch, but if you've studied it, and this improves conformance with
testsuites, I presume it's ok.
// Martin
_______________________________________________
Mingw-w64-public mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/mingw-w64-public
