Nikita Popov
cc539138ac
[CodeGen] Use __extendhfsf2 and __truncsfhf2 by default ( #126880 )
...
The standard libcalls for half to float and float to half conversion are
__extendhfsf2 and __truncsfhf2. However, LLVM currently uses
__gnu_h2f_ieee and __gnu_f2h_ieee instead. As far as I can tell, these
libcalls are an ARM-ism and only provided by libgcc on that platform.
compiler-rt always provides both libcalls.
Use the standard libcalls by default, and only use the __gnu libcalls on
ARM.
2025-02-19 10:16:57 +01:00
..
2025-02-06 18:55:50 +00:00
2025-02-06 12:41:10 -08:00
2025-01-13 14:40:25 +01:00
2025-02-13 12:13:39 +01:00
2025-02-12 07:33:22 -08:00
2025-01-28 15:55:45 +00:00
2025-02-12 14:08:18 -08:00
2024-12-16 11:37:50 +00:00
2025-02-17 12:30:07 +00:00
2024-11-26 09:13:34 -05:00
2025-02-14 09:28:57 +01:00
2025-01-23 18:23:05 +09:00
2025-02-14 09:28:57 +01:00
2025-01-28 15:18:10 -08:00
2025-02-13 10:46:42 +00:00
2025-02-06 08:58:35 +00:00
2025-01-24 10:53:11 +00:00
2024-12-16 10:49:08 +09:00
2024-12-11 17:19:07 +09:00
2025-02-13 10:46:42 +00:00
2024-11-18 18:32:20 -05:00
2025-02-12 14:08:18 -08:00
2025-02-12 14:08:18 -08:00
2025-01-24 13:27:56 +00:00
2025-02-06 14:51:13 -08:00
2025-02-06 14:51:13 -08:00
2025-01-23 18:23:05 +09:00
2025-01-28 14:02:24 +00:00
2025-02-14 12:38:04 +01:00
2025-02-05 11:41:26 +05:30
2025-02-05 15:30:33 -08:00
2025-02-14 12:44:20 +01:00
2024-12-18 06:36:22 -05:00
2024-11-02 16:54:06 -07:00
2024-11-13 09:55:37 -08:00
2024-11-13 15:35:03 -08:00
2025-02-06 14:51:13 -08:00
2024-11-02 16:54:06 -07:00
2024-11-26 09:13:34 -05:00
2024-11-22 19:48:34 +01:00
2024-12-02 16:15:52 -08:00
2024-12-04 17:17:58 +01:00
2025-01-10 22:12:47 -08:00
2025-01-17 17:55:54 -08:00
2024-11-02 16:54:06 -07:00
2025-01-23 13:46:26 -08:00
2025-02-19 10:16:57 +01:00
2025-01-02 20:13:18 -08:00
2024-11-02 16:54:06 -07:00
2024-12-03 21:33:03 -08:00
2025-01-21 00:32:56 +09:00
2025-01-30 20:58:38 +00:00
2025-01-09 09:27:29 +00:00
2025-01-27 20:30:45 +00:00
2024-12-18 09:46:45 +00:00