On Wed, 30 Jan 2019 at 10:13, Herbert Xu <herb...@gondor.apana.org.au> wrote: > > On Wed, Jan 30, 2019 at 09:33:57AM +0100, Ard Biesheuvel wrote: > > > > So what remains is the way these implementations are encapsulated by > > the crct10dif() library function, which is raster nasty, making > > CRC-T10DIF an excellent use case to discuss whether we can make any > > improvements to address some of the concerns that were also raised in > > the zinc discussion. I threw some code together a while ago [0] (and > > posted it as well, IIRC). In the mean time, a 'static call' > > infrastructure is being proposed that could be used in a similar way > > to avoid function pointers. I'm also interested in hearing opinions on > > whether the indirect call overhead is actually significant in use > > cases such as this one. > > I think even if the overhead wasn't significant it would still make > sense to make the move just for the sake of simplicity. >
Agreed, we should simplify this if we can. However, my question is whether in this particular case, a simple indirect call via a function pointer is /so/ much worse than a direct call that relies on code patching techniques that are different on every arch (and may rely on objtool or GCC plugins) that the extra complexity is justified.
