Date: Sat, 27 Jun 2026 08:23:59 -0700 From: Adrian Chadd <adrian@freebsd.org> To: Harry Schmalzbauer <freebsd@omnilan.de> Cc: ShengYi Hung <aokblast@freebsd.org>, src-committers@freebsd.org, dev-commits-src-all@freebsd.org, dev-commits-src-main@freebsd.org Subject: Re: hwpstate_intel i386 build workaround [Was: Re: git: 7b26353a59d6 - main - hwpstate_intel: Disable package control on hybrid CPU] Message-ID: <CAJ-VmonM_OxUoJgS0Fb94cgFj7dxBptcyOkBZb-RsrzTQwktFQ@mail.gmail.com> In-Reply-To: <21a39777-c38e-4805-abdc-b1c19183bce1@omnilan.de> References: <6a1e7b3b.1a45a.91f6820@gitrepo.freebsd.org> <21a39777-c38e-4805-abdc-b1c19183bce1@omnilan.de>
index | next in thread | previous in thread | raw e-mail
please create a review for this? It sounds like they should've wrapped it if #ifdef amd64 or something -a On Sat, 27 Jun 2026 at 07:19, Harry Schmalzbauer <freebsd@omnilan.de> wrote: > > On 2026-06-02 08:42, ShengYi Hung wrote: > > The branch main has been updated by aokblast: > > > > URL: https://cgit.FreeBSD.org/src/commit/?id=7b26353a59d66dc1bc611fd042a49b9e3bd13699 > > > > commit 7b26353a59d66dc1bc611fd042a49b9e3bd13699 > > Author: ShengYi Hung <aokblast@FreeBSD.org> > > AuthorDate: 2026-06-01 09:46:37 +0000 > > Commit: ShengYi Hung <aokblast@FreeBSD.org> > > CommitDate: 2026-06-02 06:41:41 +0000 > > > > hwpstate_intel: Disable package control on hybrid CPU > > > > In package control mode, the performance of all cores depends on the > > most recent value written to the request field. If the last write comes > > from an E-core, all cores are forced to align with the E-core > > performance level, resulting in significant performance degradation. > > Therefore, package control is disabled on hybrid-core systems. > > > > Reviewed by: olce > > MFC after: 2 weeks > > Sponsored by: The FreeBSD Foundation > > Sponsored by: Framework Computer Inc > > Differential Revision: https://reviews.freebsd.org/D57377 > > --- > > sys/x86/cpufreq/hwpstate_intel.c | 21 +++++++++++++++++++++ > > > In lieu of a proper fix, due to lacking skills, I'm working around i386 > incompatibility with this diff: > > diff --git a/sys/x86/cpufreq/hwpstate_intel.c > b/sys/x86/cpufreq/hwpstate_intel.c > index db8600d7b89a..b9b68f1b14d0 100644 > --- a/sys/x86/cpufreq/hwpstate_intel.c > +++ b/sys/x86/cpufreq/hwpstate_intel.c > @@ -321,6 +321,7 @@ sysctl_epp_select(SYSCTL_HANDLER_ARGS) > return (ret); > } > > +#ifndef __i386__ > static void > intel_hwpstate_hybrid_cb(void *ctx) > { > @@ -328,11 +329,14 @@ intel_hwpstate_hybrid_cb(void *ctx) > > atomic_add_32(small_cores, PCPU_GET(small_core)); > } > +#endif > > void > intel_hwpstate_identify(driver_t *driver, device_t parent) > { > +#ifndef __i386__ > uint32_t small_cores = 0; > +#endif > > if (device_find_child(parent, "hwpstate_intel", > DEVICE_UNIT_ANY) != NULL) > return; > @@ -353,6 +357,7 @@ intel_hwpstate_identify(driver_t *driver, device_t > parent) > if ((cpu_power_eax & CPUTPM1_HWP) == 0) > return; > > +#ifndef __i386__ > /* > * On hybrid-core systems, package-level control cannot be used. > * It may cause all cores to run at the E-core frequency because > @@ -363,6 +368,7 @@ intel_hwpstate_identify(driver_t *driver, device_t > parent) > intel_hwpstate_hybrid_cb, smp_no_rendezvous_barrier, > &small_cores); > if (small_cores > 0 && small_cores < mp_ncores) > hwpstate_pkg_ctrl_enable = false; > +#endif > > if (BUS_ADD_CHILD(parent, 10, "hwpstate_intel", > device_get_unit(parent)) > == NULL) > > I know i386 is not supported anymore. > Just in case anybody else wants to keep i368-stable/15 for bhyve guests > for example, where cpufreq(4) isn't attaching anyways. > > Thanks, > > -harry > >home | help
Want to link to this message? Use this
URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CAJ-VmonM_OxUoJgS0Fb94cgFj7dxBptcyOkBZb-RsrzTQwktFQ>