[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <Zxuq2Zvpn7ap4ZR5@andrea>
Date: Fri, 25 Oct 2024 17:27:37 +0300
From: Andrea Parri <parri.andrea@...il.com>
To: Hernan Ponce de Leon <hernan.poncedeleon@...weicloud.com>
Cc: puranjay@...nel.org, paulmck@...nel.org, bpf@...r.kernel.org,
lkmm@...ts.linux.dev, linux-kernel@...r.kernel.org
Subject: Re: Some observations (results) on BPF acquire and release
> I am particularly interested in tests using lwarx and stwcx instructions
> (this is what I understood would be used if one follows [1] to compile the
> tests in this thread).
>
> I have not yet check the cambridge website, but due to the timeline, I don't
> expect to find tests with those instructions. The same is true with [2].
>
> I have limited experience with diy7, but I remember that it had some
> limitations to generate RMW instructions, at least for C [3].
Oh, I'm sure there are, though I'd also not consider myself the 'expert'
when it comes to diy7 internals. ;-) Here's an example use of diy7 /
diyone7 generating lwarx and stwcx and reflecting the previous pattern:
$ diyone7 -arch PPC LwSyncdWW Coe SyncdWRPA SyncdRRAP Fre
PPC A
"LwSyncdWW Coe SyncdWRNaA SyncdRRANa Fre"
Generator=diyone7 (version 7.57+1)
Prefetch=0:x=F,0:y=W,1:y=F,1:x=T
Com=Co Fr
Orig=LwSyncdWW Coe SyncdWRNaA SyncdRRANa Fre
{
0:r2=x; 0:r4=y;
1:r2=y; 1:r3=z; 1:r6=x;
}
P0 | P1 ;
li r1,1 | li r1,2 ;
stw r1,0(r2) | stw r1,0(r2) ;
lwsync | sync ;
li r3,1 | Loop00: ;
stw r3,0(r4) | lwarx r4,r0,r3 ;
| stwcx. r4,r0,r3 ;
| bne Loop00 ;
| sync ;
| lwz r5,0(r6) ;
exists ([y]=2 /\ 1:r5=0)
But again, I'd probably have to defer to proper herdtools7 developers
and maintainers for any diy7 bug or misbehavior you'd have to discover.
Andrea
>
> Hernan
>
> [1] https://github.com/torvalds/linux/blob/master/arch/powerpc/net/bpf_jit_comp32.c
> [2] https://github.com/herd/herdtools7/tree/master/catalogue/herding-cats/ppc/tests/campaign
> [3] https://github.com/herd/herdtools7/issues/905
>
Powered by blists - more mailing lists