[PATCH v1 0/3] Fixed bugs in pl080

[PATCH v1 0/3] Fixed bugs in pl080

[1774596582]

From: TaoDing <1774596582@qq.com>

Hi all,
When I tested pl080 using the existing dmatest tool in Linux, 
I found that pl080 could not output the results correctly.
The main issues involved are inconsistent source and destination bit widths, 
misalignment of LLI, which can lead to incorrect data and even cause qemu crashes.

TaoDing (3):
  An interrupt is generated after the DMA transfer complete.
  The swidth and dwidth of pl080 are not equal.
  The LLI of pl080 should be aligned with 4 bytes.

 hw/dma/pl080.c | 37 ++++++++++++++++++++++++-------------
 1 file changed, 24 insertions(+), 13 deletions(-)

-- 
2.43.0

Re: [PATCH v1 0/3] Fixed bugs in pl080

[Tao Tang]

Hi,

It looks like three patches of this series were sent out-of-thread and 
are not attached to this cover letter.

You may want to resend the full series, or resend those patches with 
In-Reply-To pointing to the Message-ID of this cover letter so they 
appear in the same thread:

git 
send-email --in-reply-to="tencent_1BF621A658E043C088A2681178D57ED14A09@qq.com" 
xxxxx

Thanks for working on this!

Tao

Re: [PATCH v1 0/3] Fixed bugs in pl080

[Peter Maydell]

On Wed, 11 Mar 2026 at 16:14, Tao Tang <tangtao1634@phytium.com.cn> wrote:
>
> Hi,
>
> It looks like three patches of this series were sent out-of-thread and
> are not attached to this cover letter.
>
> You may want to resend the full series, or resend those patches with
> In-Reply-To pointing to the Message-ID of this cover letter so they
> appear in the same thread:
>
> git
> send-email --in-reply-to="tencent_1BF621A658E043C088A2681178D57ED14A09@qq.com"
> xxxxx

It looks to me looking at the headers that the emails were sent with correct
message-id and in-reply-to headers, but something in the qq.com
mail infrastructure has re-written the message-IDs and broken the
threading.

-- PMM

Re: [PATCH v1 0/3] Fixed bugs in pl080

[有点]

[-- Attachment #1: Type: text/plain, Size: 1817 bytes --]

Hi,


Thank you very much for your guidance
This is my first time using this system to submit code.
The message ID for patch v1 0/3 has unexpectedly changed, I am trying to submit it using a other email.




         原始邮件
         
       
发件人：Peter Maydell <peter.maydell@linaro.org&gt;
发件时间：2026年3月12日 01:05
收件人：Tao Tang <tangtao1634@phytium.com.cn&gt;
抄送：1774596582 <1774596582@qq.com&gt;, qemu-arm <qemu-arm@nongnu.org&gt;, qemu-devel <qemu-devel@nongnu.org&gt;
主题：Re: [PATCH v1 0/3] Fixed bugs in pl080



       On&nbsp;Wed,&nbsp;11&nbsp;Mar&nbsp;2026&nbsp;at&nbsp;16:14,&nbsp;Tao&nbsp;Tang&nbsp;<tangtao1634@phytium.com.cn&gt;&nbsp;wrote:
&gt;
&gt;&nbsp;Hi,
&gt;
&gt;&nbsp;It&nbsp;looks&nbsp;like&nbsp;three&nbsp;patches&nbsp;of&nbsp;this&nbsp;series&nbsp;were&nbsp;sent&nbsp;out-of-thread&nbsp;and
&gt;&nbsp;are&nbsp;not&nbsp;attached&nbsp;to&nbsp;this&nbsp;cover&nbsp;letter.
&gt;
&gt;&nbsp;You&nbsp;may&nbsp;want&nbsp;to&nbsp;resend&nbsp;the&nbsp;full&nbsp;series,&nbsp;or&nbsp;resend&nbsp;those&nbsp;patches&nbsp;with
&gt;&nbsp;In-Reply-To&nbsp;pointing&nbsp;to&nbsp;the&nbsp;Message-ID&nbsp;of&nbsp;this&nbsp;cover&nbsp;letter&nbsp;so&nbsp;they
&gt;&nbsp;appear&nbsp;in&nbsp;the&nbsp;same&nbsp;thread:
&gt;
&gt;&nbsp;git
&gt;&nbsp;send-email&nbsp;--in-reply-to="tencent_1BF621A658E043C088A2681178D57ED14A09@qq.com"
&gt;&nbsp;xxxxx

It&nbsp;looks&nbsp;to&nbsp;me&nbsp;looking&nbsp;at&nbsp;the&nbsp;headers&nbsp;that&nbsp;the&nbsp;emails&nbsp;were&nbsp;sent&nbsp;with&nbsp;correct
message-id&nbsp;and&nbsp;in-reply-to&nbsp;headers,&nbsp;but&nbsp;something&nbsp;in&nbsp;the&nbsp;qq.com
mail&nbsp;infrastructure&nbsp;has&nbsp;re-written&nbsp;the&nbsp;message-IDs&nbsp;and&nbsp;broken&nbsp;the
threading.

--&nbsp;PMM

[-- Attachment #2: Type: text/html, Size: 4197 bytes --]

[PATCH v1 0/3] Fixed bugs in pl080

[Tao Ding]

Hi all,
When I tested pl080 using the existing dmatest tool in Linux, 
I found that pl080 could not output the results correctly.
The main issues involved are inconsistent source and destination bit widths, 
misalignment of LLI, which can lead to incorrect data and even cause qemu crashes.

Tao Ding (3):
  An interrupt is generated after the DMA transfer complete.
  The swidth and dwidth of pl080 are not equal.
  The LLI of pl080 should be aligned with 4 bytes.

 hw/dma/pl080.c | 45 ++++++++++++++++++++++++++++-----------------
 1 file changed, 28 insertions(+), 17 deletions(-)


base-commit: 1fd5ff9d76d23ab23a68419cbc76d5ee33e8b455
-- 
2.43.0

Re: [PATCH v1 0/3] Fixed bugs in pl080

[Peter Maydell]

On Thu, 12 Mar 2026 at 08:02, Tao Ding <dingtao0430@163.com> wrote:
>
> Hi all,
> When I tested pl080 using the existing dmatest tool in Linux,
> I found that pl080 could not output the results correctly.
> The main issues involved are inconsistent source and destination bit widths,
> misalignment of LLI, which can lead to incorrect data and even cause qemu crashes.
>
> Tao Ding (3):
>   An interrupt is generated after the DMA transfer complete.
>   The swidth and dwidth of pl080 are not equal.
>   The LLI of pl080 should be aligned with 4 bytes.

Hi; I've taken patches 1 and 3 into target-arm.next (with
some adjustments to the commit messages). I had some
review comments on patch 2.

thanks
-- PMM