Date: Fri, 26 Mar 2021 20:17:42 +0100 From: Andrea Venturoli <ml@netfence.it> To: freebsd-hackers@freebsd.org Subject: Re: cost/benefit of some src.conf options Message-ID: <79a899c3-368e-20d2-8ac7-0741e00fa3b1@netfence.it> In-Reply-To: <56F46324-59BB-4CC2-BE90-5FF63C4554ED@FreeBSD.org> References: <YF3pHo5Pj5Swm90O@ceres.zyxst.net> <56F46324-59BB-4CC2-BE90-5FF63C4554ED@FreeBSD.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On 3/26/21 7:46 PM, Dimitry Andric wrote: > That said, the retpoline mechanisms tend to be fairly well tested by > now, but will still have a non-negligible performance impact, maybe even > a large impact, depending on your workload. There is no simple answer > here, you will have to measure it for yourself. AFAIK: _ RETpoline is an alternative to IBRS; _ the impact of RETpoline should be lower than IBRS; _ IBRS is enabled by default. Did I get it wrong? So, unless someone is willing to disable IBRS and live without mitigation, it would be interesting to know how performance differs between the two. I've seen IBRS's impact on bhyve-hosted Windows guests reach 15%-20%. I've never tried RETpoline for the lack of information WRT to its stability: I guess "fairly well tested" does not mean "production ready", or it would be enabled by default, wouldn't it? :)
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?79a899c3-368e-20d2-8ac7-0741e00fa3b1>