From: "Sireesh Kodali" <sireeshkodali1@gmail.com>
To: "Vinod Koul" <vkoul@kernel.org>
Cc: <linux-kernel@vger.kernel.org>, <linux-doc@vger.kernel.org>,
<linux-arm-msm@vger.kernel.org>, <dmaengine@vger.kernel.org>,
<~postmarketos/upstreaming@lists.sr.ht>
Subject: Re: [PATCH v2 0/3] dmaengine: Add support for immediate commands
Date: Fri, 25 Nov 2022 22:20:18 +0530 [thread overview]
Message-ID: <COLJ587CFY67.2E0WQZD08PLBY@skynet-linux> (raw)
In-Reply-To: <Y3FudBqc1vQ8fEgU@matsya>
On Mon Nov 14, 2022 at 3:53 AM IST, Vinod Koul wrote:
> On 11-11-22, 10:42, Sireesh Kodali wrote:
> > On Fri Nov 4, 2022 at 6:10 PM IST, Vinod Koul wrote:
> > > On 27-10-22, 10:44, Sireesh Kodali wrote:
> > > > The IPA v2.x block, found on some older Qualcomm SoCs, uses BAM DMA to
> > > > send and receive packets from the AP. It also uses BAM to receive
> > > > commands from the AP (and possibly the modem). These commands are
> > > > encoded as "Immediate Commands". They vary from regular BAM DMA
> > > > commands. Adding support for immediate commands is trivial, but requires
> > > > also adding Immediate Commands to the dmaengine API, which is what this
> > > > patch series does.
> > >
> > > Can you explain a bit more. I understand you need "Immediate Commands"
> > > but am really reluctant to add another interface to support a specific
> > > use case
> > >
> >
> > Apologies for the delayed response
> >
> > BAM supports both regular commands, and "immediate commands". Currently,
> > commands are used by the Qualcom NAND chip driver, while "immediate
> > commands" are intended to be used by the (yet to be mainlined) IPA
> > driver. From the BAM driver perspective, both immediate and regular
> > commands are simply a matter of setting the appropriate flag in the
> > descriptor. I don't have access to the documentation on BAM to know
> > exactly how these two modes differ, however I do know they are not
> > interchangable. If a different API is suggested, I can change the
> > implementation as needed.
>
> Ok, can you please explain what is meant by 'regular' cmd and
> 'immediate', lets see what is required here
Stephan pointed out the APQ8016E TRM has details on BAM. As I understand
it, 'regular' commands are queued register read/writes for the
peripheral. Immediate commands on the other hand seem to be interpreted
by the peripheral's firmware, and don't involve any register
writes/reads from BAM's perspective.
This is what the TRM has to say:
> Immediate (IMM) (only for BAM-NDP): Allows the software to create
> descriptors of type immediate, which does not generate any data
> transmissions or registers configuration, it is simply supplied to the
> peripheral, the peripheral then parses its fields (which are
> irrelevant to the BAM). Only the flags of this descriptor are relevant
> to the BAM, address and size are irrelevant, and BAM simply passes
> them as is to the peripheral. This can be used for the software to
> operate peripheral-specific operations within regular data operations.
> Immediate descriptors are published on the sidebands as 1 byte size
> descriptor, once BAM_NDP fetches an immediate descriptor, it publishes
> all recently fetched descriptors including the immediate descriptor
> with immediate indication, to inform the peripheral that the last
> published descriptor was immediate descriptor.
> Command (CMD) (only for BAM-lite and BAM-NDP): Allows the software to
> create descriptors of type command. Descriptors of type command do not
> generate any data transmissions but configure registers in the
> peripheral (write and read registers operations)
Regards,
Sireesh
>
> --
> ~Vinod
prev parent reply other threads:[~2022-11-25 16:51 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-10-27 5:14 [PATCH v2 0/3] dmaengine: Add support for immediate commands Sireesh Kodali
2022-10-27 5:14 ` [PATCH v2 1/3] doc: dmaengine: client-api: Add immediate commands in the DMA client API Sireesh Kodali
2022-10-27 5:14 ` [PATCH v2 2/3] dmaengine: Add support for immediate commands in the " Sireesh Kodali
2022-10-27 5:14 ` [PATCH v2 3/3] dmaengine: bam_dma: Add support for immediate commands Sireesh Kodali
2022-11-04 12:40 ` [PATCH v2 0/3] dmaengine: " Vinod Koul
2022-11-11 5:12 ` Sireesh Kodali
2022-11-13 22:23 ` Vinod Koul
2022-11-24 3:14 ` Sireesh Kodali
2022-11-25 16:50 ` Sireesh Kodali [this message]
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=COLJ587CFY67.2E0WQZD08PLBY@skynet-linux \
--to=sireeshkodali1@gmail.com \
--cc=dmaengine@vger.kernel.org \
--cc=linux-arm-msm@vger.kernel.org \
--cc=linux-doc@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=vkoul@kernel.org \
--cc=~postmarketos/upstreaming@lists.sr.ht \
/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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.