3 Commits

Author	SHA1	Message	Date
Andrew Stubbs	12e3bac3ce	amdgcn: remove unnecessary scalar cache flush ... The exit code isn't actually written via the scalar cache so the cache flush is not actually needed.	2023-10-31 13:36:45 +01:00
Tobias Burnus	b9898fc993	amdgcn: Replace asm("s8") by __builtin_gcn_kernarg_ptr if existing ... Check whether __builtin_gcn_kernarg_ptr is available and, if it is, call it instead using the hard-coded 'asm("s8")' in: * newlib/libc/machine/amdgcn/exit-value.h (exit_with_int) * newlib/libc/machine/amdgcn/mlock.c (sbrk) * newlib/libc/sys/amdgcn/write.c (write) newlib/libc/machine/amdgcn/exit-value.h | 6 ++++++ newlib/libc/machine/amdgcn/mlock.c | 10 +++++++--- newlib/libc/sys/amdgcn/write.c | 4 ++++ 3 files changed, 17 insertions(+), 3 deletions(-)	2022-11-21 13:10:29 +01:00
Jeff Johnston	1787e9d033	AMD GCN Port contributed by Andrew Stubbs <ams@codesourcery.com> ... Add support for the AMD GCN GPU architecture. This is primarily intended for use with OpenMP and OpenACC offloading. It can also be used for stand-alone programs, but this is intended mostly for testing the compiler and is not expected to be useful in general. The GPU architecture is highly parallel, and therefore Newlib must be configured to use dynamic re-entrancy, and thread-safe malloc. The only I/O available is a via a shared-memory interface provided by libgomp and the gcn-run tool included with GCC. At this time this is limited to stdout, argc/argv, and the return code.	2019-01-15 10:48:08 -05:00