Hi, On Thu Mar 20, 2025 at 2:54 PM CET, Jason Merrill wrote:
[...] >>>>> Thanks a lot. Merged to trunk as r15-7238-gceabea405ffdc8. >>>>> >>>>> Even though it’s an “old” regression (from GCC 7), it’s a >>>>> reject-valid and I’m tempted to backport; OK for active branches as >>>>> well? >>>> >>>> For backports I'd be inclined to condition the INTEGER_TYPE handing on >>>> if (processing_template_decl) to reduce its impact. But first let's >>>> wait and see if this breaks anything on trunk. >>> ACK. I’ll resume the discussion about backports via this thread in 2-3 >>> weeks. >> Please find attached the updated patch for active branches with the >> restriction you suggested. >> >> Successfully tested on x86_64-pc-linux-gnu. OK for both GCC 13 and 14? > > OK. Thanks. Backported as r13-9441-g65e998d172e006cdf0dd4d58f83784a5fed61fc5 and r14-11426-gf86d274ab76fdd89d7afd9b2eab28f3a61749cfb. Simon
