Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
FROMLIST: ARM: NWFPE: avoid compiler-generated __aeabi_uldivmod
clang-15's ability to elide loops completely became more aggressive when it can deduce how a variable is being updated in a loop. Counting down one variable by an increment of another can be replaced by a modulo operation. For 64b variables on 32b ARM EABI targets, this can result in the compiler generating calls to __aeabi_uldivmod, which it does for a do while loop in float64_rem(). For the kernel, we'd generally prefer that developers not open code 64b division via binary / operators and instead use the more explicit helpers from div64.h. On arm-linux-gnuabi targets, failure to do so can result in linkage failures due to undefined references to __aeabi_uldivmod(). While developers can avoid open coding divisions on 64b variables, the compiler doesn't know that the Linux kernel has a partial implementation of a compiler runtime (--rtlib) to enforce this convention. It's also undecidable for the compiler whether the code in question would be faster to execute the loop vs elide it and do the 64b division. While I actively avoid using the internal -mllvm command line flags, I think we get better code than using barrier() here, which will force reloads+spills in the loop for all toolchains. Link: ClangBuiltLinux/linux#1666 Reported-by: Nathan Chancellor <nathan@kernel.org> Reviewed-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Nick Desaulniers <ndesaulniers@google.com> Bug: 250930332 Link: https://lore.kernel.org/lkml/20221010225342.3903590-1-ndesaulniers@google.com/ Change-Id: I038f06e43420e72927956133a8fe39a3db21a9db
- Loading branch information