Date: Fri, 20 Oct 2017 14:40:21 +0000 From: "Bjoern A. Zeeb" <bz@zabbadoz.net> To: "Ryan Stone" <rysto32@gmail.com> Cc: "Bret Ketchum" <bcketchum@gmail.com>, "Ruslan Bukin" <br@FreeBSD.org>, "freebsd-arch@freebsd.org" <freebsd-arch@freebsd.org> Subject: Re: PEBS support in hwpmc Message-ID: <BC911E8F-A88E-41B6-8048-287C8D31E596@zabbadoz.net> In-Reply-To: <CAFMmRNyf0vDiU52kLzV01=1qrhT1Zzv6-AJCfDv1kGzH1Hwm=A@mail.gmail.com> References: <CAGm6yaRGeX0yG2OdntVQh4gvzbxhG6QboEufhw-Xj=a_UNnTHg@mail.gmail.com> <CAFMmRNyf0vDiU52kLzV01=1qrhT1Zzv6-AJCfDv1kGzH1Hwm=A@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On 20 Oct 2017, at 14:24, Ryan Stone wrote: > On Fri, Oct 20, 2017 at 7:27 AM, Bret Ketchum <bcketchum@gmail.com> > wrote: >> Without this support (or a VTune subscription) understanding >> Front-End/Back-End bound applications running on Skylake/Kaby Lake >> processors will be difficult at best > > I'm afraid that I don't know of any work related to PEBS in hwpmc. > However, I'm curious as to why PEBS is so important on these > architectures. My experience with hwpmc profiling has been that > callchain information is frequently critical for understanding the > performance characteristics and my understanding is that PEBS by > design cannot capture that information. Ruslan is working on Intel PT to my best knowledge. Cc:ing him so he can chime in. /bz
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?BC911E8F-A88E-41B6-8048-287C8D31E596>