[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <a6a150f332b44914ac286550b63feab0@AcuMS.aculab.com>
Date: Fri, 31 Mar 2023 15:59:44 +0000
From: David Laight <David.Laight@...LAB.COM>
To: "'Kirill A. Shutemov'" <kirill@...temov.name>,
Tom Lendacky <thomas.lendacky@....com>
CC: 'Wu Zongyong' <wuzongyong@...ux.alibaba.com>,
"kirill.shutemov@...ux.intel.com" <kirill.shutemov@...ux.intel.com>,
"tglx@...utronix.de" <tglx@...utronix.de>,
"mingo@...hat.com" <mingo@...hat.com>,
"dave.hansen@...ux.intel.com" <dave.hansen@...ux.intel.com>,
"x86@...nel.org" <x86@...nel.org>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
"tony.luck@...el.com" <tony.luck@...el.com>,
"wutu.xq2@...ux.alibaba.com" <wutu.xq2@...ux.alibaba.com>
Subject: RE: [RFC PATCH] x86/insn: support decode MOVSXD instruction for MMIO
From: Kirill A. Shutemov
> Sent: 31 March 2023 16:25
...
> > No, SEV doesn't support userspace MMIO.
>
> But where do you filter out userspace MMIO? AFAICS, it goes straight from
> from #VC to insn_decode_mmio(). Hm?
Probably by making vm_iomap_memory() fail.
Otherwise MOVSXD is the least of your problems.
You'd need to worry about all the AVX opcodes as well.
Although you might even find kernel code that is using
kernel_fpu_begin/end() to wrap mmio copies that use the
big AVX512 registers.
When each PCIe read takes about 1us (measured into our fpga)
increasing the TLP to 64 bytes (from 8) makes a massive
difference to buffer reads.
(Mostly we try to get the fpga to do writes instead.)
David
-
Registered Address Lakeside, Bramley Road, Mount Farm, Milton Keynes, MK1 1PT, UK
Registration No: 1397386 (Wales)
Powered by blists - more mailing lists