MIME-Version: 1.0
References: <20230303231133.1486085-1-eranian@google.com> <20230306120106.GE1267364@hirez.programming.kicks-ass.net>
In-Reply-To: <20230306120106.GE1267364@hirez.programming.kicks-ass.net>
From:   Nick Desaulniers <ndesaulniers@google.com>
Date:   Mon, 6 Mar 2023 16:16:52 -0800
Message-ID: <CAKwvOdnRvd5KK01awAyeyt5S36TPPW4_8Z6YL1r4gB-pBrHTbg@mail.gmail.com>
Subject: Re: [PATCH] x86/resctrl: avoid compiler optimization in __resctrl_sched_in
To:     Peter Zijlstra <peterz@infradead.org>
Cc:     Stephane Eranian <eranian@google.com>,
        linux-kernel@vger.kernel.org, tony.luck@intel.com,
        reinette.chatre@intel.com, fenghua.yu@intel.com,
        peternewman@google.com, james.morse@arm.com, babu.moger@amd.com,
        ananth.narayan@amd.com, vschneid@redhat.com,
        Nathan Chancellor <nathan@kernel.org>,
        clang-built-linux <llvm@lists.linux.dev>,
        Borislav Petkov <bp@alien8.de>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

Start of Lore thread:
https://lore.kernel.org/lkml/20230303231133.1486085-1-eranian@google.com/

On Mon, Mar 6, 2023 at 4:01=E2=80=AFAM Peter Zijlstra <peterz@infradead.org=
> wrote:
>
> On Fri, Mar 03, 2023 at 03:11:33PM -0800, Stephane Eranian wrote:
>
> > The problem is located in the __resctrl_sched_in() routine which rewrit=
es
> > the active closid via the PQR_ASSOC register. Because this is an expens=
ive
> > operation, the kernel only does it when the context switch involves tas=
ks
> > with different CLOSID. And to check that, it needs to access the curren=
t
> > task's closid field using current->closid. current is actually a macro
> > that reads the per-cpu variable pcpu_hot.current_task.
> >
> > After an investigation by compiler experts, the problem has been tracke=
d down
> > to the usage of the get_current() macro in the __resctrl_sched_in() cod=
e and
> > in particular the per-cpu macro:
> >
> > static __always_inline struct task_struct *get_current(void)
> > {
> >         return this_cpu_read_stable(pcpu_hot.current_task);
> > }
> >
> > And as per percpu.h:
> >
> > /*
> >  * this_cpu_read() makes gcc load the percpu variable every time it is
> >  * accessed while this_cpu_read_stable() allows the value to be cached.
> >  * this_cpu_read_stable() is more efficient and can be used if its valu=
e
> >  * is guaranteed to be valid across cpus.  The current users include
> >  * get_current() and get_thread_info() both of which are actually
> >  * per-thread variables implemented as per-cpu variables and thus
> >  * stable for the duration of the respective task.
> >  */
> >
> > The _stable version of the macro allows the value to be cached, meaning=
 it
> > does not force a reload.
>
> Right, so afaict the difference between this_cpu_read() and
> this_cpu_read_stable() is the volatile qualifier.
>
> this_cpu_read() is asm volatile(), while this_cpu_read_stable() and
> raw_cpu_read() are both an unqualified asm().
>
> Now, afaiu we're inlining all of this into __switch_to(), which has
> raw_cpu_write(pcpu_hot.current_task, next_p).
>
> And I suppose what the compiler is doing is lifting the 'current' load
> over that store, but how is it allowed that? I thought C was supposed to
> have PO consistency, That raw_cpu_write() should be seen as a store to
> to pcpu_hot.current_task, why can it lift a load over the store?
>
> Specifically, percpu_to_op() has a "+m" output constaint while
> percpu_stable_op() has a "p" input constraint on the same address.

I definitely think the issue is specific to "p" constraints.
https://godbolt.org/z/34YeG6WbY is the test case I reduced which I
think demonstrates the issue.

https://reviews.llvm.org/D145416
-> click "Show Older Changes" for the ongoing discussion.

I don't have a satisfactory answer yet, but am looking into this.

>
> Compiler folks help?


--=20
Thanks,
~Nick Desaulniers