Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Date:   Wed, 19 May 2021 21:22:11 +0200
From:   Peter Zijlstra <peterz@infradead.org>
To:     Thomas Gleixner <tglx@linutronix.de>
Cc:     Naresh Kamboju <naresh.kamboju@linaro.org>,
        LKML <linux-kernel@vger.kernel.org>,
        linux-mm <linux-mm@kvack.org>, lkft-triage@lists.linaro.org,
        Andrew Morton <akpm@linux-foundation.org>,
        Ira Weiny <ira.weiny@intel.com>,
        Oliver Sang <oliver.sang@intel.com>,
        Chaitanya Kulkarni <Chaitanya.Kulkarni@wdc.com>,
        x86@kernel.org, Kalesh Singh <kaleshsingh@google.com>
Subject: Re: BUG: unable to handle page fault for address - EIP:
 __kmap_local_page_prot
Message-ID: <20210519192211.GI21560@worktop.programming.kicks-ass.net>
References: <CA+G9fYtAbUGO9oAtL8eZ9Pu-_a1wx3y8Tk=pDO3Fh3dEwoRGWg@mail.gmail.com>
 <87mtsqd0c1.ffs@nanos.tec.linutronix.de>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <87mtsqd0c1.ffs@nanos.tec.linutronix.de>
Precedence: bulk

On Wed, May 19, 2021 at 08:50:54PM +0200, Thomas Gleixner wrote:
> On Wed, May 19 2021 at 20:38, Naresh Kamboju wrote:
> > While running LTP mm test suite on i386 kernel the following warning and BUG
> > reported on linux next 5.13.0-rc2-next-20210519.
> >
> > The warning is not regression, We have been noticing these warnings often on
> > i386
> 
> Noticing? Did you report them?
> 
> > but kernel BUG: looks to be a new crash.
> 
> It's just the consequence of the problem which is pointed out by the
> warning. Fix below.

Maybe something like so as well?

---
 mm/mremap.c | 8 ++++++++
 1 file changed, 8 insertions(+)

diff --git a/mm/mremap.c b/mm/mremap.c
index 47c255b60150..477a908bdb6b 100644
--- a/mm/mremap.c
+++ b/mm/mremap.c
@@ -210,6 +210,10 @@ static void move_ptes(struct vm_area_struct *vma, pmd_t *old_pmd,
 }
 
 #ifdef CONFIG_HAVE_MOVE_PMD
+#if CONFIG_PGTABLE_LEVELS < 2
+#error "Can't very well move non-existant pgtable levels"
+#endif
+
 static bool move_normal_pmd(struct vm_area_struct *vma, unsigned long old_addr,
 		  unsigned long new_addr, pmd_t *old_pmd, pmd_t *new_pmd)
 {
@@ -277,6 +281,10 @@ static inline bool move_normal_pmd(struct vm_area_struct *vma,
 #endif
 
 #ifdef CONFIG_HAVE_MOVE_PUD
+#if CONFIG_PGTABLE_LEVELS < 3
+#error "Can't very well move non-existant pgtable levels"
+#endif
+
 static bool move_normal_pud(struct vm_area_struct *vma, unsigned long old_addr,
 		  unsigned long new_addr, pud_t *old_pud, pud_t *new_pud)
 {