Re: [PATCH] x86-64: Add --enable-x86-64-mfentry

2025-07-11 Thread Siddhesh Poyarekar
On 2025-07-11 15:28, Uros Bizjak wrote: Why not just switch over unconditionally? __fentry__ seems like a better alternative to mcount overall and it has been around long enough that even older deployments should be relatively unaffected. Actually, it is switched on by default for i?86-*-linux

Re: [PATCH] x86-64: Add --enable-x86-64-mfentry

2025-07-11 Thread Sam James
Uros Bizjak writes: > On Fri, Jul 11, 2025 at 2:33 PM Siddhesh Poyarekar > wrote: >> >> On 2025-07-08 18:07, Sam James wrote: >> >> OK in principle, but please allow some time for distro maintainers >> >> (CC'd) to voice their opinion. >> > >> > It looks good to me and I plan on us using it. I'

Re: [PATCH] x86-64: Add --enable-x86-64-mfentry

2025-07-11 Thread Uros Bizjak
On Fri, Jul 11, 2025 at 2:33 PM Siddhesh Poyarekar wrote: > > On 2025-07-08 18:07, Sam James wrote: > >> OK in principle, but please allow some time for distro maintainers > >> (CC'd) to voice their opinion. > > > > It looks good to me and I plan on us using it. I'd like opinions from > > one othe

Re: [PATCH] x86-64: Add --enable-x86-64-mfentry

2025-07-11 Thread Sam James
Siddhesh Poyarekar writes: > On 2025-07-08 18:07, Sam James wrote: >>> OK in principle, but please allow some time for distro maintainers >>> (CC'd) to voice their opinion. >> It looks good to me and I plan on us using it. I'd like opinions >> from >> one other group first before it goes in if po

Re: [PATCH] x86-64: Add --enable-x86-64-mfentry

2025-07-11 Thread Siddhesh Poyarekar
On 2025-07-08 18:07, Sam James wrote: OK in principle, but please allow some time for distro maintainers (CC'd) to voice their opinion. It looks good to me and I plan on us using it. I'd like opinions from one other group first before it goes in if possible though, as our perspective is differe

Re: [PATCH] x86-64: Add --enable-x86-64-mfentry

2025-07-08 Thread Sam James
Uros Bizjak writes: > On Thu, Jul 3, 2025 at 12:01 PM H.J. Lu wrote: >> >> When profiling is enabled with shrink wrapping, the mcount call may not >> be placed at the function entry after >> >> pushq %rbp >> movq %rsp,%rbp >> >> As the result, the profile data may be skewed which makes PGO less

Re: [PATCH] x86-64: Add --enable-x86-64-mfentry

2025-07-03 Thread Uros Bizjak
On Thu, Jul 3, 2025 at 12:01 PM H.J. Lu wrote: > > When profiling is enabled with shrink wrapping, the mcount call may not > be placed at the function entry after > > pushq %rbp > movq %rsp,%rbp > > As the result, the profile data may be skewed which makes PGO less > effective. > > Add --enable-x8