maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity

Fix this C99 DR to be more robust 

This should fix the following test issue on ARM:
https://lab.llvm.org/buildbot/#/builders/171/builds/16815
This commit is contained in:
Aaron Ballman 2022-07-01 07:33:37 -04:00
parent 583abd0e36
commit 14035d5147
1 changed files with 11 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

15
clang/test/C/drs/dr2xx.c
View File

@ -31,16 +31,23 @@ void dr204(void) {
/* If the implementation supports a standard integer type larger than signed
* long, it's okay for size_t and ptrdiff_t to have a greater integer
* conversion rank than signed long.
*
* Note, it's not required that the implementation use that larger conversion
* rank; it's acceptable to use an unsigned long or unsigned int for the size
* type (those ranks are not greater than that of signed long).
*/
(void)_Generic(s + sl, __typeof__(s) : 1);
(void)_Generic(p + sl, __typeof__(p) : 1);
(void)_Generic(s + sl, __typeof__(s) : 1, unsigned long : 1, unsigned int : 1);
(void)_Generic(p + sl, __typeof__(p) : 1, signed long : 1, signed int : 1);
#elif __LLONG_WIDTH__ == __LONG_WIDTH__
/* But if the implementation doesn't support a larger standard integer type
* than signed long, the conversion rank should prefer signed long if the type
* is signed (ptrdiff_t) or unsigned long if the type is unsigned (size_t).
*
* Note, as above, unsigned/signed int is also acceptable due to having a
* lesser integer conversion rank.
*/
(void)_Generic(s + sl, unsigned long : 1);
(void)_Generic(p + sl, signed long : 1);
(void)_Generic(s + sl, unsigned long : 1, unsigned int : 1);
(void)_Generic(p + sl, signed long : 1, signed int : 1);
#else
#error "Something has gone off the rails"
#endif