Hi Jin: Curious, did you introduce version info into the multi-lib path in your donwstream GCC? I suspect we need a more complete solution to deal with the multiversion issue, so it would be great if you can provide more experience with this part?
Thanks :) On Wed, Feb 22, 2023 at 5:44 PM Jin Ma <ji...@linux.alibaba.com> wrote: > > When there is an extension with different versions, the result of the > TARGET_COMPUTE_MULTILIB hook > is generally wrong, so the version needs to be considered. > > gcc/ChangeLog: > > * common/config/riscv/riscv-common.cc > (riscv_subset_list::match_score): Match the version > of each extension. > --- > gcc/common/config/riscv/riscv-common.cc | 5 ++--- > 1 file changed, 2 insertions(+), 3 deletions(-) > > diff --git a/gcc/common/config/riscv/riscv-common.cc > b/gcc/common/config/riscv/riscv-common.cc > index ebc1ed7d7e4..0a15ff705d1 100644 > --- a/gcc/common/config/riscv/riscv-common.cc > +++ b/gcc/common/config/riscv/riscv-common.cc > @@ -331,10 +331,9 @@ riscv_subset_list::match_score (riscv_subset_list *list) > const > /* list must be subset of current this list, otherwise it not safe to > link. > TODO: We might give different weight for each extension, but the rule > could > - be complicated. > - TODO: We might consider the version of each extension. */ > + be complicated. */ > for (s = list->m_head; s != NULL; s = s->next) > - if (this->lookup (s->name.c_str ()) != NULL) > + if (this->lookup (s->name.c_str (), s->major_version, s->minor_version) > != NULL) > score++; > else > return 0; > -- > 2.17.1 >
