From: "André Przywara" <andre.przywara@arm.com>
To: Amit Tomer <amittomer25@gmail.com>
Cc: vkoul@kernel.org, "Andreas Färber" <afaerber@suse.de>,
"Manivannan Sadhasivam" <manivannan.sadhasivam@linaro.org>,
dan.j.williams@intel.com, cristian.ciocaltea@gmail.com,
dmaengine@vger.kernel.org,
linux-arm-kernel <linux-arm-kernel@lists.infradead.org>,
linux-actions@lists.infradead.org
Subject: Re: [PATCH RFC 2/8] dmaengine: Actions: Add support for S700 DMA engine
Date: Wed, 6 May 2020 14:04:05 +0100 [thread overview]
Message-ID: <f4791ad3-1edb-3f24-5bcd-b2859e4a1742@arm.com> (raw)
In-Reply-To: <CABHD4K9mqpcO7jo4NQov__8jEGhAJr2o8JTiX1N+Z=zb9vG0OQ@mail.gmail.com>
On 06/05/2020 13:54, Amit Tomer wrote:
> Hi,
>
> Thanks for quick review
>
>> You should mention (at least in the commit message) why this is needed.
>> And please move this into a separate function, this indentation is
>> becoming mad here
>
> There is not much documented about it, and all I see is GIC crash
> if I keep it open for S700. Would figure out more details about it and
> update in next version.
What I meant is that you should mention this and the fact that you got
this bit from the BSP source.
> .
>>
>>> + for (i = 0; i < od->nr_pchans; i++) {
>>> + pchan = &od->pchans[i];
>>> + chan_irq_pending = pchan_readl(pchan,
>>> + OWL_DMAX_INT_CTL) &
>>> + pchan_readl(pchan,
>>> + OWL_DMAX_INT_STATUS)
>>> + ;
>>> +
>>> + /* Dummy read to ensure OWL_DMA_IRQ_PD0 value is
>>> + * updated
>>> + */
>>> + dma_readl(od, OWL_DMA_IRQ_PD0);
>>>
>>> - global_irq_pending = dma_readl(od, OWL_DMA_IRQ_PD0);
>>> + global_irq_pending = dma_readl(od,
>>> + OWL_DMA_IRQ_PD0);
>>>
>>> - if (chan_irq_pending && !(global_irq_pending & BIT(i))) {
>>> - dev_dbg(od->dma.dev,
>>> - "global and channel IRQ pending match err\n");
>>> + if (chan_irq_pending && !(global_irq_pending &
>>> + BIT(i))) {
>>> + dev_dbg(od->dma.dev,
>>> + "global and channel IRQ pending match err\n");
>>>
>>> - /* Clear IRQ status for this pchan */
>>> - pchan_update(pchan, OWL_DMAX_INT_STATUS,
>>> - 0xff, false);
>>> + /* Clear IRQ status for this pchan */
>>> + pchan_update(pchan, OWL_DMAX_INT_STATUS,
>>> + 0xff, false);
>>>
>>> - /* Update global IRQ pending */
>>> - pending |= BIT(i);
>>> + /* Update global IRQ pending */
>>> + pending |= BIT(i);
>>> + }
>>> }
>>> }
>>>
>>> @@ -720,6 +743,7 @@ static int owl_dma_resume(struct dma_chan *chan)
>>>
>>> static u32 owl_dma_getbytes_chan(struct owl_dma_vchan *vchan)
>>> {
>>> + struct owl_dma *od = to_owl_dma(vchan->vc.chan.device);
>>> struct owl_dma_pchan *pchan;
>>> struct owl_dma_txd *txd;
>>> struct owl_dma_lli *lli;
>>> @@ -741,9 +765,15 @@ static u32 owl_dma_getbytes_chan(struct owl_dma_vchan *vchan)
>>> list_for_each_entry(lli, &txd->lli_list, node) {
>>> /* Start from the next active node */
>>> if (lli->phys == next_lli_phy) {
>>> - list_for_each_entry(lli, &txd->lli_list, node)
>>> - bytes += lli->hw[OWL_DMADESC_FLEN] &
>>> - GENMASK(19, 0);
>>> + list_for_each_entry(lli, &txd->lli_list, node) {
>>> + if (od->devid == S700_DMA)
>>> + bytes +=
>>> + lli->hw[OWL_DMADESC_FLEN];
>>> + else
>>> + bytes +=
>>> + lli->hw[OWL_DMADESC_FLEN] &
>>> + GENMASK(19, 0);
>>
>> You should have an accessor for getting the frame len, that should avoid
>> the insane wrapping here. Or factor this out into a helper function.
>> Alternatively revert the if statement and continue, that saves you one
>> level of indentation.
>>
>> I guess flen is limited to 20 bits anyway, so you might want to apply
>> the 20-bit mask unconditionally.
>
> Actually, on S700 flen uses 24 bits , so we should not use 20-bit mask.
I think it *can* use 24 bits. Where does flen come from? I guess it is
less than 1 MB anyway already? It better should be, at least, since the
S900 seems to have this limit.
> For accessor function, shall this be okay ?
Something like that. My point was that this can be much more simplified
if you go with 20 bits *always*. Then you can save the first parameter
and this becomes a one-liner.
Cheers,
Andre
> +static u32 llc_hw_flen(struct owl_dma *od,
> + struct owl_dma_lli *lli)
> +{
> + u32 bit_mask;
> +
> + if (od->devid == S700_DMA)
> + bit_mask = 23;
> + else
> + bit_mask = 19;
> +
> + return lli->hw[OWL_DMADESC_FLEN] & GENMASK(bit_mask, 0);
> +
> +}
>
> Thanks
> Amit
>
prev parent reply other threads:[~2020-05-06 13:04 UTC|newest]
Thread overview: 12+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <1588761371-9078-1-git-send-email-amittomer25@gmail.com>
2020-05-06 10:36 ` [PATCH RFC 1/8] dmaengine: Actions: get rid of bit fields from dma descriptor Amit Singh Tomar
2020-05-10 15:51 ` Manivannan Sadhasivam
2020-05-11 10:45 ` Amit Tomer
2020-05-11 11:20 ` Manivannan Sadhasivam
2020-05-11 11:44 ` André Przywara
2020-05-11 12:04 ` Manivannan Sadhasivam
2020-05-11 12:48 ` André Przywara
2020-05-11 15:29 ` Manivannan Sadhasivam
2020-05-06 10:36 ` [PATCH RFC 2/8] dmaengine: Actions: Add support for S700 DMA engine Amit Singh Tomar
2020-05-06 11:12 ` André Przywara
2020-05-06 12:54 ` Amit Tomer
2020-05-06 13:04 ` André Przywara [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=f4791ad3-1edb-3f24-5bcd-b2859e4a1742@arm.com \
--to=andre.przywara@arm.com \
--cc=afaerber@suse.de \
--cc=amittomer25@gmail.com \
--cc=cristian.ciocaltea@gmail.com \
--cc=dan.j.williams@intel.com \
--cc=dmaengine@vger.kernel.org \
--cc=linux-actions@lists.infradead.org \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=manivannan.sadhasivam@linaro.org \
--cc=vkoul@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).