Date: Sun, 15 Feb 2009 13:59:41 -0000 From: Joseph Koshy <jkoshy@FreeBSD.org> To: Robert Watson <rwatson@FreeBSD.org> Cc: Sam, Joseph Koshy <jkoshy@FreeBSD.org>, Leffler <sam@freebsd.org>, freebsd-hackers@freebsd.org, Max Laier <max@love2party.net>, Andrew Brampton <brampton+freebsd-hackers@gmail.com> Subject: Re: pahole - Finding holes in kernel structs Message-ID: <86skmfztrm.wl%koshy@unixconsulting.co.in> In-Reply-To: <alpine.BSF.2.00.0902121752470.2205@fledge.watson.org> References: <d41814900902120608i4b54c86fp9f565bbeead5a476@mail.gmail.com> <200902121717.47841.max@love2party.net> <4994516B.8060703@freebsd.org> <200902121754.24536.max@love2party.net> <alpine.BSF.2.00.0902121752470.2205@fledge.watson.org>
next in thread | previous in thread | raw e-mail | index | archive | help
> Well, I think we want to inform this through actual measurement. Right now, > tools like hwpmc track cache misses by point in executable code, but what > would be nice is if we could post-process to generate cache miss information > by data structure field... That is one of the tools that I've had in mind. The hard part in the process is tracking sample locations back to source artifacts with acceptable accuracy. I couldn't find existing libraries with a compatible open-source license that I could build on, and thus the motivation for some of the libraries listed in the roadmap at http://elftoolchain.sourceforge.net/. Also, hwpmc(4) would need to be augmented to use "precise" sampling on the hardware that supports it, but this is relatively straightforward. Koshy
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?86skmfztrm.wl%koshy>