linux-kernel - Re: [PATCH] drm/panfrost: Don't try to map on error faults

Open Source and information security mailing list archives

Message-ID: <565bac84-9e3c-21de-9043-f508e98ae6bf@arm.com>
Date:   Wed, 5 Feb 2020 14:01:47 +0000
From:   Steven Price <steven.price@....com>
To:     Robin Murphy <robin.murphy@....com>,
        Tomeu Vizoso <tomeu.vizoso@...labora.com>,
        linux-kernel@...r.kernel.org
Cc:     David Airlie <airlied@...ux.ie>,
        Alyssa Rosenzweig <alyssa.rosenzweig@...labora.com>,
        dri-devel@...ts.freedesktop.org
Subject: Re: [PATCH] drm/panfrost: Don't try to map on error faults

On 05/02/2020 13:25, Robin Murphy wrote:
> On 05/02/2020 10:07 am, Tomeu Vizoso wrote:
>> If the exception type isn't one of the normal faults, don't try to map
>> and instead go straight to a terminal fault.
> 
> "One of the the normal faults" seems a rather vague way of saying "a
> translation fault", which is what we're specifically handling here, and
> logically the only fault reflecting something not yet mapped rather than
> mapped inappropriately ;)
> 
> (Who knows how the level ended up as 1-4 rather than 0-3 as it really
> should be - another Mali Mystery(TM)...)

Ah, but you're thinking of LPAE not Mali Magic Page Tables ('inspired
by' LPAE but not actually compatible....) ;) However I do wonder what
will happen when we enable aarch64 page tables in Bifrost.

>> Otherwise, we can get flooded by kernel warnings and further faults.
> 
> Either way,
> 
> Reviewed-by: Robin Murphy <robin.murphy@....com>
> 
>> Signed-off-by: Tomeu Vizoso <tomeu.vizoso@...labora.com>
>> ---
>>   drivers/gpu/drm/panfrost/panfrost_mmu.c | 5 +++--
>>   1 file changed, 3 insertions(+), 2 deletions(-)
>>
>> diff --git a/drivers/gpu/drm/panfrost/panfrost_mmu.c
>> b/drivers/gpu/drm/panfrost/panfrost_mmu.c
>> index 763cfca886a7..80abddb4544c 100644
>> --- a/drivers/gpu/drm/panfrost/panfrost_mmu.c
>> +++ b/drivers/gpu/drm/panfrost/panfrost_mmu.c
>> @@ -596,8 +596,9 @@ static irqreturn_t
>> panfrost_mmu_irq_handler_thread(int irq, void *data)
>>           source_id = (fault_status >> 16);
>>             /* Page fault only */
>> -        if ((status & mask) == BIT(i)) {
>> -            WARN_ON(exception_type < 0xC1 || exception_type > 0xC4);
>> +        if ((status & mask) == BIT(i) &&
>> +             exception_type >= 0xC1 &&
>> +             exception_type <= 0xC4) {

I would suggest the best option here is to copy mali_kbase and check
against a mask, in kbase[1] we have:

	switch (fault_status & AS_FAULTSTATUS_EXCEPTION_CODE_MASK) {

	case AS_FAULTSTATUS_EXCEPTION_CODE_TRANSLATION_FAULT:

[1]
https://gitlab.freedesktop.org/panfrost/linux-panfrost/blob/a864e6b9fbd093a033a90d5bbdd6dcc79a0667b2/mali_kbase_mmu.c#L108

Where:

#define AS_FAULTSTATUS_EXCEPTION_CODE_MASK                      (0x7<<3)
#define AS_FAULTSTATUS_EXCEPTION_CODE_TRANSLATION_FAULT         (0x0<<3)

Steve

>>                 ret = panfrost_mmu_map_fault_addr(pfdev, i, addr);
>>               if (!ret) {
>>
> _______________________________________________
> dri-devel mailing list
> dri-devel@...ts.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/dri-devel

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives