Re: [PATCH 5.9 24/74] x86, powerpc: Rename memcpy_mcsafe() to copy_mc_to_{user, kernel}()

[Hans-Peter Jansen]

Hi Greg, hi Dan,

Am Samstag, 31. Oktober 2020, 12:36:06 CET schrieb Greg Kroah-Hartman:
> From: Dan Williams <dan.j.williams@xxxxxxxxx>
> 
> commit ec6347bb43395cb92126788a1a5b25302543f815 upstream.
> 
> In reaction to a proposal to introduce a memcpy_mcsafe_fast()
> implementation Linus points out that memcpy_mcsafe() is poorly named
> relative to communicating the scope of the interface. Specifically what
> addresses are valid to pass as source, destination, and what faults /
> exceptions are handled.
> 
> 
> Introduce an x86 copy_mc_fragile() name as the rename for the
> low-level x86 implementation formerly named memcpy_mcsafe(). It is used
> as the slow / careful backend that is supplanted by a fast
> copy_mc_generic() in a follow-on patch.
> 
> One side-effect of this reorganization is that separating copy_mc_64.S
> to its own file means that perf no longer needs to track dependencies
> for its memcpy_64.S benchmarks.
> 
> ---
> arch/powerpc/lib/copy_mc_64.S                          |  242 +++++++++++++++++
> arch/powerpc/lib/memcpy_mcsafe_64.S                    |  242 -----------------

> tools/testing/selftests/powerpc/copyloops/copy_mc_64.S |  242 +++++++++++++++++ 

This change leaves a dangling symlink in 
tools/testing/selftests/powerpc/copyloops behind. At least, this is, what I 
could track down, when building 5.9.3 within an environment, that bails out 
on this:

[ 2908s] calling /usr/lib/rpm/brp-suse.d/brp-25-symlink
[ 2908s] ERROR: link target doesn't exist (neither in build root nor in installed system):
[ 2908s]   /usr/src/linux-5.9.3-lp152.3-vanilla/tools/testing/selftests/powerpc/copyloops/memcpy_mcsafe_64.S -> /usr/src/linux-5.9.3-lp152.3-vanilla/arch/powerpc/lib/memcpy_mcsafe_64.S
[ 2908s] Add the package providing the target to BuildRequires and Requires
[ 2909s] INFO: relinking /usr/src/linux-5.9.3-lp152.3-vanilla/tools/testing/selftests/powerpc/primitives/asm/asm-compat.h -> ../../../../../../arch/powerpc/include/asm/asm-compat.h (was ../.././../../../../arch/powerpc/include/asm/asm-compat.h)

Linus` tree seems to not suffer from this, though.

Cheers,
Pete