Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
DMARC-Filter: OpenDMARC Filter v1.3.2 smtp.codeaurora.org 72A2F61834
Subject: Re: [PATCH] dmaengine: qcom-bam: fix circular buffer handling
To:     Srinivas Kandagatla <srinivas.kandagatla@linaro.org>,
        vkoul@kernel.org
Cc:     dmaengine@vger.kernel.org, linux-kernel@vger.kernel.org,
        linux-arm-msm@vger.kernel.org
References: <20190614142012.31384-1-srinivas.kandagatla@linaro.org>
 <f4522b78-b406-954c-57b7-923e6ab31f96@codeaurora.org>
 <d84af3ad-5ba4-0f24-fd30-2fa20cf85658@linaro.org>
 <2d370a33-fa16-45ca-cf82-9d775349f806@codeaurora.org>
 <544851f6-58b8-2506-01ce-5c4d1f93fb3c@linaro.org>
 <a50066ac-be85-6706-e7f3-f1069fd0dd0b@codeaurora.org>
 <31574ef2-d675-bb36-08d1-18b756ebd29e@linaro.org>
From:   Sricharan R <sricharan@codeaurora.org>
Message-ID: <d86b1c3b-19bc-6c51-a364-c46ca019db1a@codeaurora.org>
Date:   Wed, 19 Jun 2019 23:20:05 +0530
User-Agent: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:60.0) Gecko/20100101
 Thunderbird/60.7.0
MIME-Version: 1.0
In-Reply-To: <31574ef2-d675-bb36-08d1-18b756ebd29e@linaro.org>
Content-Type: text/plain; charset=utf-8
Content-Language: en-US
Content-Transfer-Encoding: 8bit
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

Hi Srini,

On 6/18/2019 10:20 PM, Srinivas Kandagatla wrote:
> 
> 
> On 18/06/2019 17:27, Sricharan R wrote:
>>   The Macro's expect that buffer size is power of 2. So we are infact passing the actual correct
>>   size ( MAX_DESCRIPTORS + 1 = 4096)
> This will make the circular buffer macros happy but question is that do we actually have that many descriptor buffers?
> 
> This is what is in the driver:
> 
> #define BAM_DESC_FIFO_SIZE    SZ_32K
> #define MAX_DESCRIPTORS (BAM_DESC_FIFO_SIZE / sizeof(struct bam_desc_hw) - 1)
> #define BAM_FIFO_SIZE    (SZ_32K - 8)
> 
> Wouldn't having MAX_DESCRIPTORS + 1 = 4096  lead to overflow the actual descriptor memory size of (SZ_32K - 8) ?
> 

Right, but the CIRC_SPACE macro assumes there is 1 space less than the actual size.
That said, agree there is an issue on the boundary. I will also do some testing tomorrow
on this and get back.

Regards,
 Sricharan
-- 
"QUALCOMM INDIA, on behalf of Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, hosted by The Linux Foundation