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;
Subject: Re: [PATCH 4/4] DMA: XILINX_ZYNQMP_DPDMA depends on HAS_IOMEM
To:     Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Cc:     linux-kernel@vger.kernel.org, Stefan Roese <sr@denx.de>,
        Vinod Koul <vkoul@kernel.org>, dmaengine@vger.kernel.org,
        Sinan Kaya <okaya@codeaurora.org>,
        Green Wan <green.wan@sifive.com>,
        Hyun Kwon <hyun.kwon@xilinx.com>,
        Tejas Upadhyay <tejasu@xilinx.com>,
        Michal Simek <michal.simek@xilinx.com>,
        kernel test robot <lkp@intel.com>
References: <20210522021313.16405-1-rdunlap@infradead.org>
 <20210522021313.16405-5-rdunlap@infradead.org>
 <YKmfs68Cq4osBaQ0@pendragon.ideasonboard.com>
From:   Randy Dunlap <rdunlap@infradead.org>
Message-ID: <5cb3b313-cd96-d687-2916-0d4af8e5e675@infradead.org>
Date:   Sat, 22 May 2021 18:07:01 -0700
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:78.0) Gecko/20100101
 Thunderbird/78.10.0
MIME-Version: 1.0
In-Reply-To: <YKmfs68Cq4osBaQ0@pendragon.ideasonboard.com>
Content-Type: text/plain; charset=utf-8
Content-Language: en-US
Content-Transfer-Encoding: 7bit
Precedence: bulk

On 5/22/21 5:20 PM, Laurent Pinchart wrote:
> Hi Randy,
> 
> Thank you for the patch.
> 
> On Fri, May 21, 2021 at 07:13:13PM -0700, Randy Dunlap wrote:
>> When CONFIG_HAS_IOMEM is not set/enabled, most iomap() family
>> functions [including ioremap(), devm_ioremap(), etc.] are not
>> available.
>> Drivers that use these functions should depend on HAS_IOMEM so that
>> they do not cause build errors.
>>
>> Cures this build error:
>> s390-linux-ld: drivers/dma/xilinx/xilinx_dpdma.o: in function `xilinx_dpdma_probe':
>> xilinx_dpdma.c:(.text+0x336a): undefined reference to `devm_platform_ioremap_resource'
> 
> I've previously posted
> https://lore.kernel.org/dmaengine/20210520152420.23986-2-laurent.pinchart@ideasonboard.com/T/#u
> which fixes the same issue (plus an additional one).

Hi Laurent,

I didn't add a dependency on OF because OF header files _mostly_
have stubs so that they work when  OF is enabled or disabled.

I did find a problem in <linux/of_address.h> where it could end up
without having a stub. I will post a patch for that soon.
I'm currently doing lots of randconfig builds on it.

Thanks.

>> Fixes: 7cbb0c63de3fc ("dmaengine: xilinx: dpdma: Add the Xilinx DisplayPort DMA engine driver")
>> Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
>> Reported-by: kernel test robot <lkp@intel.com>
>> Cc: Vinod Koul <vkoul@kernel.org>
>> CC: dmaengine@vger.kernel.org
>> Cc: Hyun Kwon <hyun.kwon@xilinx.com>
>> Cc: Tejas Upadhyay <tejasu@xilinx.com>
>> Cc: Michal Simek <michal.simek@xilinx.com>
>> Cc: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
>> ---
>>  drivers/dma/Kconfig |    1 +
>>  1 file changed, 1 insertion(+)
>>
>> --- linux-next-20210521.orig/drivers/dma/Kconfig
>> +++ linux-next-20210521/drivers/dma/Kconfig
>> @@ -702,6 +702,7 @@ config XILINX_ZYNQMP_DMA
>>  
>>  config XILINX_ZYNQMP_DPDMA
>>  	tristate "Xilinx DPDMA Engine"
>> +	depends on HAS_IOMEM
>>  	select DMA_ENGINE
>>  	select DMA_VIRTUAL_CHANNELS
>>  	help
> 


-- 
~Randy