Age | Commit message (Collapse) | Author | Files | Lines |
|
|
|
Since they are laways paired there is no need to "search" for the cuda
linker.
|
|
Fixes issue #5870
|
|
class
Instead of the DynamicLinker returning a hardcoded value like
`-Wl,-foo`, it now is passed a value that could be '-Wl,', or could be
something '-Xlinker='
This makes a few things cleaner, and will make it possible to fix using
clang (not clang-cl) on windows, where it invokes either link.exe or
lld-link.exe instead of a gnu-ld compatible linker.
|
|
Most of the cuda code is from Olexa Bilaniuk.
Most of the PGI code is from Michael Hirsc
|
|
|
|
Even with the check (for extra safety) string.replace is more than twice
as fast.
|
|
I debated a bit whether both classes really belong in the same module,
and decided that they do because the share a number of helpers.
|
|
In most cases instead pass `for_machine`, the name of the relevant
machines (what compilers target, what targets run on, etc). This allows
us to use the cross code path in the native case, deduplicating the
code.
As one can see, environment got bigger as more information is kept
structured there, while ninjabackend got a smaller. Overall a few amount
of lines were added, but the hope is what's added is a lot simpler than
what's removed.
|
|
|
|
|
|
Partially addresses #4961.
|
|
PArtially addresses #4961.
Also make the sanity check do double duty as a GPU architecture
detection test.
|
|
|
|
Also adds test case for shared library linking.
Closes #4912, at least on Linux. The future 0.50.0 does not yet claim to
support CUDA on systems other than Linux and backends other than Ninja.
|
|
|
|
|
|
|
|
|
|
|