From: Marek Vasut <marek.vasut@mailbox.org>
To: Frank Li <Frank.li@nxp.com>
Cc: linux-media@vger.kernel.org, Fabio Estevam <festevam@gmail.com>,
Laurent Pinchart <laurent.pinchart@ideasonboard.com>,
Mauro Carvalho Chehab <mchehab@kernel.org>,
Ming Qian <ming.qian@oss.nxp.com>,
Mirela Rabulea <mirela.rabulea@nxp.com>,
Nicolas Dufresne <nicolas.dufresne@collabora.com>,
Pengutronix Kernel Team <kernel@pengutronix.de>,
Sascha Hauer <s.hauer@pengutronix.de>,
Shawn Guo <shawnguo@kernel.org>,
imx@lists.linux.dev, linux-arm-kernel@lists.infradead.org
Subject: Re: [PATCH] media: imx-jpeg: Add support for descriptor allocation from SRAM
Date: Wed, 20 Aug 2025 22:16:09 +0200 [thread overview]
Message-ID: <7439c678-5369-4c31-9c89-5a015ce7582f@mailbox.org> (raw)
In-Reply-To: <aKYRJap9Gwc2mFXh@lizhi-Precision-Tower-5810>
On 8/20/25 8:17 PM, Frank Li wrote:
> On Wed, Aug 20, 2025 at 06:29:53PM +0200, Marek Vasut wrote:
>> Add support for optional allocation of bitstream descriptors from SRAM
>> instead of DRAM. In case the encoder/decoder DT node contains 'sram'
>> property which points to 'mmio-sram', the driver will attempt to use
>> the SRAM instead of DRAM for descriptor allocation, which might improve
>> performance.
>>
>> This however helps on i.MX95 with sporadic SLOTn_STATUS IMG_RD_ERR bit 11
>> being triggered during JPEG encoding. The following pipeline triggers the
>> problem when descriptors get allocated from DRAM, the pipeline often hangs
>> after a few seconds and the encoder driver indicates "timeout, cancel it" :
>
> Do you know why this happen if descriptor is in DRAM? why sram help this
> case?
No. Did NXP ever observe this ?
It seems to me the encode never received much testing, considering the
severe bugs that showed up in there.
>> gst-launch-1.0 videotestsrc ! video/x-raw,width=256,height=256,format=YUY2 ! \
>> queue ! v4l2jpegenc ! queue ! fakesink
>>
>> Signed-off-by: Marek Vasut <marek.vasut@mailbox.org>
>> ---
> ...
>> - GFP_ATOMIC);
>> + jpeg->slot_data.cfg_dec_vaddr = mxc_jpeg_alloc(jpeg, jpeg->slot_data.cfg_dec_size,
>> + &jpeg->slot_data.cfg_dec_daddr);
>> if (!jpeg->slot_data.cfg_dec_vaddr)
>> goto err;
>>
>> @@ -2902,6 +2911,10 @@ static int mxc_jpeg_probe(struct platform_device *pdev)
>> jpeg->dev = dev;
>> jpeg->mode = mode;
>>
>> + /* SRAM pool is optional */
>> + jpeg->sram_pool = of_gen_pool_get(pdev->dev.of_node, "sram", 0);
>
> Need update binding doc nxp,imx8-jpeg.yaml
Sure
next prev parent reply other threads:[~2025-08-21 1:04 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-08-20 16:29 [PATCH] media: imx-jpeg: Add support for descriptor allocation from SRAM Marek Vasut
2025-08-20 18:17 ` Frank Li
2025-08-20 20:16 ` Marek Vasut [this message]
2025-08-21 6:28 ` Ming Qian(OSS)
2025-08-21 9:03 ` Marek Vasut
2025-08-21 9:42 ` Ming Qian(OSS)
2025-08-21 12:01 ` Marek Vasut
2025-08-22 0:58 ` Ming Qian(OSS)
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=7439c678-5369-4c31-9c89-5a015ce7582f@mailbox.org \
--to=marek.vasut@mailbox.org \
--cc=Frank.li@nxp.com \
--cc=festevam@gmail.com \
--cc=imx@lists.linux.dev \
--cc=kernel@pengutronix.de \
--cc=laurent.pinchart@ideasonboard.com \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-media@vger.kernel.org \
--cc=mchehab@kernel.org \
--cc=ming.qian@oss.nxp.com \
--cc=mirela.rabulea@nxp.com \
--cc=nicolas.dufresne@collabora.com \
--cc=s.hauer@pengutronix.de \
--cc=shawnguo@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).