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: <87mvpfgjk2.fsf@intel.com>
Date:	Thu, 31 Mar 2016 11:04:13 +0300
From:	Felipe Balbi <felipe.balbi@...ux.intel.com>
To:	"Thang Q. Nguyen" <tqnguyen@....com>
Cc:	Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
	linux-usb@...r.kernel.org, linux-omap@...r.kernel.org,
	linux-kernel@...r.kernel.org, linux-arm@...ts.infradead.org,
	Phong Vo <pvo@....com>, Loc Ho <lho@....com>,
	patches <patches@....com>
Subject: Re: [PATCH v3 1/2] usb:dwc3: Enable support for 64-bit system


Hi,

(please don't top-post)

"Thang Q. Nguyen" <tqnguyen@....com> writes:
> Hi Balbi,
> If CONFIG_DMA_CMA=y, dma mask is set properly. The issue just happen
> when CONFIG_DMA_CMA is not set. In this case, dma mask is not set and
> we need this code to check if dma mask should be manually set to 32 or
> 64.

Can you point me to the code which has this conditional ? Why would
DMA_CMA=n mean that dma_mask isn't initialized ? According to DMA_CMA's
help text (see below) that's supposed to allow drivers to *allocate*
large contiguous buffers, but that's not the case here.

config DMA_CMA
	bool "DMA Contiguous Memory Allocator"
	depends on HAVE_DMA_CONTIGUOUS && CMA
	help
	  This enables the Contiguous Memory Allocator which allows drivers
	  to allocate big physically-contiguous blocks of memory for use with
	  hardware components that do not support I/O map nor scatter-gather.

	  You can disable CMA by specifying "cma=0" on the kernel's command
	  line.

	  For more information see <include/linux/dma-contiguous.h>.
	  If unsure, say "n".



>
> ----
> Thang
>
> On Wed, Mar 30, 2016 at 8:09 PM, Felipe Balbi
> <felipe.balbi@...ux.intel.com> wrote:
>>
>> Hi,
>>
>> "Thang Q. Nguyen" <tqnguyen@....com> writes:
>>> From: "Thang Q. Nguyen" <tqnguyen@....com>
>>>
>>> Add 64-bit DMA operation support to the USB DWC3 driver.
>>> First attempt to set the coherent DMA mask for 64-bit DMA.
>>> If that failed, attempt again with 32-bit DMA.
>>>
>>> Changes from v2:
>>>       - None.
>>>
>>> Changes from v1:
>>>       - Remove WARN_ON if dma_mask is NULL
>>
>> these changes lines should be between the tearline (---) and diffstat
>> below.
>>
>>> Signed-off-by: Thang Q. Nguyen <tqnguyen@....com>
>>> ---
>>>  drivers/usb/dwc3/core.c | 15 +++++++++++++++
>>>  1 file changed, 15 insertions(+)
>>>
>>> diff --git a/drivers/usb/dwc3/core.c b/drivers/usb/dwc3/core.c
>>> index de5e01f..2479c24 100644
>>> --- a/drivers/usb/dwc3/core.c
>>> +++ b/drivers/usb/dwc3/core.c
>>> @@ -831,6 +831,21 @@ static int dwc3_probe(struct platform_device *pdev)
>>>       dwc->mem = mem;
>>>       dwc->dev = dev;
>>>
>>> +     /* Try to set 64-bit DMA first */
>>> +     if (!pdev->dev.dma_mask)
>>> +             /* Platform did not initialize dma_mask */
>>> +             ret = dma_coerce_mask_and_coherent(&pdev->dev,
>>> +                                                DMA_BIT_MASK(64));
>>> +     else
>>> +             ret = dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(64));
>>> +
>>> +     /* If seting 64-bit DMA mask fails, fall back to 32-bit DMA mask */
>>> +     if (ret) {
>>> +             ret = dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(32));
>>> +             if (ret)
>>> +                     return ret;
>>> +     }
>>
>> Also, why is it so that you need this now ? glue layers are copying dma
>> mask from parent device and that should be set properly. This really
>> shouldn't be necessary in dwc3-core; it would mean that glue layer
>> didn't set this device up properly.
>>
>> --
>> balbi
>
>
>
> -- 
>
> Thang Q. Nguyen    | Staff SW Eng.
>
> C: +849.7684.7606 | O: +848.3770.0640
>
> F: +848.3770.0641  | tqnguyen@....com

-- 
balbi

Download attachment "signature.asc" of type "application/pgp-signature" (819 bytes)

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ