linux-kernel - Re: Compilation error report for: drivers/firmware/qcom_scm.c:469:47: error: passing argument 3 of ‘dma_alloc

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

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <CAK8P3a2UQ2xNs6R+YC8owEWQ-ooH=tuERa7Fs7acnkwi3N=FeA@mail.gmail.com>
Date:   Tue, 30 Jan 2018 15:17:54 +0100
From:   Arnd Bergmann <arnd@...db.de>
To:     Benjamin Gaignard <benjamin.gaignard@...aro.org>
Cc:     Benjamin GAIGNARD <benjamin.gaignard@...com>,
        Arnaud POULIQUEN <arnaud.pouliquen@...com>,
        "Andy Gross <andy.gross@...aro.org> , David Brown" 
        <david.brown@...aro.org>,
        "linux-arm-msm@...r.kernel.org" <linux-arm-msm@...r.kernel.org>,
        "linux-soc@...r.kernel.org" <linux-soc@...r.kernel.org>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: Compilation error report for: drivers/firmware/qcom_scm.c:469:47: error: passing argument 3 of ‘dma_alloc_coherent’ from incompatible pointer type

On Tue, Jan 30, 2018 at 3:00 PM, Benjamin Gaignard
<benjamin.gaignard@...aro.org> wrote:
> 2018-01-30 14:25 GMT+01:00 Arnd Bergmann <arnd@...db.de>:
>> On Tue, Jan 30, 2018 at 11:11 AM, Benjamin GAIGNARD
>> <benjamin.gaignard@...com> wrote:
>>>
>>> On 01/12/2018 05:11 PM, Arnaud Pouliquen wrote:
>>>> Hello Andy,David,
>>> + Arnd
>>>
>>> I have the same issue on drm-misc-next.
>>> Does Arnaud's fix make sense or should we update/change the way of how
>>> we compile the kernel ?
>>
>> We've hit a couple of bugs with qcom drivers confusing physical addresses
>> and DMA addresses in the past, usually the drivers were buggy in
>> some form, and tried to use dma_alloc_coherent() to get a buffer
>> that gets passed into a firmware interface taking a physical address,
>> which is of course completely wrong.
>>
>>>> diff --git a/drivers/firmware/qcom_scm.c b/drivers/firmware/qcom_scm.c
>>>> index af4c752..8dfbe61 100644
>>>> --- a/drivers/firmware/qcom_scm.c
>>>> +++ b/drivers/firmware/qcom_scm.c
>>>> @@ -448,7 +448,7 @@ int qcom_scm_assign_mem(phys_addr_t mem_addr, size_t
>>>> mem_sz,
>>>>       struct qcom_scm_mem_map_info *mem_to_map;
>>>>       phys_addr_t mem_to_map_phys;
>>>>       phys_addr_t dest_phys;
>>>> -     phys_addr_t ptr_phys;
>>>> +     dma_addr_t ptr_phys;
>>>>       size_t mem_to_map_sz;
>>>>       size_t dest_sz;
>>>>       size_t src_sz;
>>
>> This would be bad: you can basically never have a 'dma_addr_t ptr_phys': it can
>> be exactly one of 'dma address',  'physical address' or a pointer,
>> this claims that the
>> struct member is all three of them.
>>
>> The proper fix here is to stop using dma_alloc_coherent.
>
> Okay but that doesn't explain why we are the only ones to get an issue
> while the parameter
> doesn't match function prototype

For almost all configurations, dma_addr_t and phys_addr_t are the
same width, and gcc treats them as compatible.

I usually get the warning during randconfig builds, but you seem to
have started with a configuration like this.

        Arnd