[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <055f4c2b-0e7f-44ae-92ff-a1025a217208@citrix.com>
Date: Fri, 15 Aug 2025 12:19:31 +0100
From: Andrew Cooper <andrew.cooper3@...rix.com>
To: Peter Zijlstra <peterz@...radead.org>
Cc: "H. Peter Anvin" <hpa@...or.com>, x86@...nel.org, kees@...nel.org,
alyssa.milburn@...el.com, scott.d.constable@...el.com,
joao@...rdrivepizza.com, samitolvanen@...gle.com, nathan@...nel.org,
alexei.starovoitov@...il.com, mhiramat@...nel.org, ojeda@...nel.org,
linux-kernel@...r.kernel.org
Subject: Re: [RFC][PATCH] x86,ibt: Use UDB instead of 0xEA
On 15/08/2025 11:59 am, Peter Zijlstra wrote:
> On Fri, Aug 15, 2025 at 11:43:11AM +0100, Andrew Cooper wrote:
>
>>>> I've had to use cs prefixed jcc.d32, because our older toolchains don't
>>>> like the ,pn notation.
>>> And then I forgot to move that cs prefix around in the bhi1 case...
>>> fixed that.
>> Dare I ask what ,pn notation is? It's not only the older toolchains
>> that don't know it :)
> Some fancy notation for the CS/DS branch hints.
>
> CS Jcc, decodes to Jcc,pn for non-taken
> DS Jcc, decodes to Jcc,pt for taken
Ah, thanks. I was looking at the hex in one of the comments and still
couldn't figure it out.
So with this notation, we also have the dual meaning of ,pt between the
P4 and LNC. At least the encoding is the same.
~Andrew
Powered by blists - more mailing lists