Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Fri, 8 Jul 2022 13:48:46 -0700 (PDT)
From:   Hugh Dickins <hughd@google.com>
To:     Nadav Amit <namit@vmware.com>
cc:     Hugh Dickins <hughd@google.com>,
        LKML <linux-kernel@vger.kernel.org>,
        Thomas Gleixner <tglx@linutronix.de>,
        Ingo Molnar <mingo@redhat.com>, Borislav Petkov <bp@alien8.de>,
        "x86@kernel.org" <x86@kernel.org>, Linux MM <linux-mm@kvack.org>,
        Dave Hansen <dave.hansen@linux.intel.com>,
        Peter Zijlstra <peterz@infradead.org>,
        Andy Lutomirski <luto@kernel.org>,
        Andrew Morton <akpm@linux-foundation.org>
Subject: Re: [PATCH] x86/mm/tlb: ignore f->new_tlb_gen when zero
In-Reply-To: <49C173B5-CF12-44AA-BE00-689E99E65C23@vmware.com>
Message-ID: <23b7e5f8-8b38-51aa-fecd-3ed92118af72@google.com>
References: <20220708003053.158480-1-namit@vmware.com> <4eccbb3a-2b5a-039-e934-da451e469929@google.com> <49C173B5-CF12-44AA-BE00-689E99E65C23@vmware.com>
MIME-Version: 1.0
Content-Type: multipart/mixed; boundary="0-2142999139-1657313342=:1568"
Precedence: bulk

  This message is in MIME format.  The first part should be readable text,
  while the remaining parts are likely unreadable without MIME-aware tools.

--0-2142999139-1657313342=:1568
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: QUOTED-PRINTABLE

On Fri, 8 Jul 2022, Nadav Amit wrote:
> On Jul 8, 2022, at 12:21 PM, Hugh Dickins <hughd@google.com> wrote:
>=20
> > =E2=9A=A0 External Email
> >=20
> > On Thu, 7 Jul 2022, Nadav Amit wrote:
> >=20
> >> From: Nadav Amit <namit@vmware.com>
> >>=20
> >> Commit aa44284960d5 ("x86/mm/tlb: Avoid reading mm_tlb_gen when
> >> possible") introduced an optimization of skipping the flush if the TLB
> >> generation that is flushed (as provided in flush_tlb_info) was already
> >> flushed.
> >>=20
> >> However, arch_tlbbatch_flush() does not provide any generation in
> >> flush_tlb_info. As a result, try_to_unmap_one() would not perform any
> >> TLB flushes.
> >>=20
> >> Fix it by checking whether f->new_tlb_gen is nonzero. Zero value is
> >> anyhow is an invalid generation value.
> >>=20
> >> In addition, add the missing unlikely() and jump to get tracing right.
> >>=20
> >> Fixes: aa44284960d5 ("x86/mm/tlb: Avoid reading mm_tlb_gen when possib=
le")
> >> Reported-by: Hugh Dickins <hughd@google.com>
> >> Cc: Dave Hansen <dave.hansen@linux.intel.com>
> >> Cc: Peter Zijlstra (Intel) <peterz@infradead.org>
> >> Cc: Andy Lutomirski <luto@kernel.org>
> >> Signed-off-by: Nadav Amit <namit@vmware.com>
> >=20
> > Thanks a lot for your rapid response and thinking it through
> > (before I got around to any "nopcid" or "nopti" experiments).
> >=20
> > I've been testing this one for a few hours now, and no problems seen.
> > I expect you'll be sending another version, maybe next week, meeting
> > Dave's concerns; but wanted to reassure that you have correctly
> > identified the issue and fixed it with this - thanks.
>=20
> Thanks, Hugh. Sorry again for my mistake.
>=20
> Can I please have your =E2=80=9CTested-by=E2=80=9D?

Sure, let me scrabble around in my box of tags, yes, here's one:

Tested-by: Hugh Dickins <hughd@google.com>
--0-2142999139-1657313342=:1568--