On Thu, Oct 24, 2024 at 9:48 PM David Brown via Gcc <gcc@gcc.gnu.org> wrote: > > On 24/10/2024 16:35, Jonathan Wakely via Gcc wrote: > > On Thu, 24 Oct 2024 at 15:00, Mateusz Guzik via Gcc <gcc@gcc.gnu.org> wrote: > > > >> I understand the stock behavior of pilling variables on may happen to > >> improve cache usage. > >> > >> However, in a multicore setting it is a never-ending source of > >> unintentionally showing up and disappearing false-sharing depending on > >> unrelated variables being added or removed. > >> > >> I'm aware one can manually add padding or explicitly tell the linker > >> to put a specific variable elsewhere, but in large codebases getting > >> to a point where everything is annotated is quite a challenge. To give > >> you one example of a project which did not get there despite literally > >> decades of multicore development: the Linux kernel (or in fact any > >> open-source kernel out there, last I checked). > >> > >> In the meantime a great clutch would contain such problems to a > >> specific .o file. > >> > ... > >> For illustrative purposes, I don't care about the name: > >> -align-object-data-section=64 > >> > >> Thoughts? > >> > > > > Wouldn't that be a linker option, and so not part of GCC? > > > > I would have thought it would be better as part of the compiler.
It also would need compiler support for the LTO case where what is a linker "object unit" is different from the source "translation unit". But there the question is what's the desired behavior? Currently you'd need to use -flto-partition=1to1 to get the mapping of translation unit to object unit the same as when not using LTO. Richard. > For > each compilation unit, you generate one or more data sections depending > on the variable initialisations, compiler options and target (.bss, > .data, .rodata, .sbss, etc.). If the compiler has > "-align-object-data-section=64" in effect, then you could add ".align > 64" to the start and the end of each section. As far as I can see, that > would give the effect the OP is looking for in a simple manner that can > also be different for different translation units in the same program > (which would be hard to do with a linker flag). And if this is a flag > in gcc, then it will work for any ld-compatible linker (gold, mold, etc.). > > gcc already generates a ".align" linker command with the minimum > alignment required by the variables in the section. This new flag would > then do two things - one is set a minimum alignment for the data > sections (it can still increase if a variable has bigger alignment, such > as via an "_Alignas" specifier or an "aligned" attribute). Secondly, it > would add another ".align" command at the end of the section so that if > a section with a small alignment specifier is linked after one with a > specified large alignment, you'll get padding inserted. > > mvh., > > David >
