From: nicolas.ferre@atmel.com (Nicolas Ferre)
To: linux-arm-kernel@lists.infradead.org
Subject: [PATCH v2 2/2] dmaengine: hdmac: Add scatter-gathered memset support
Date: Thu, 22 Oct 2015 12:18:51 +0200 [thread overview]
Message-ID: <5628B80B.8040105@atmel.com> (raw)
In-Reply-To: <1445506860-22619-3-git-send-email-maxime.ripard@free-electrons.com>
Le 22/10/2015 11:41, Maxime Ripard a ?crit :
> Just like memset support, the HDMAC might be used to do a memset over a
> discontiguous memory area.
>
> In such a case, we'll just build up a chain of memset descriptors over the
> contiguous chunks of memory to set, in order to allow such a support.
>
> Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
Yep, seems good!
Acked-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Thanks Maxime, bye.
> ---
> drivers/dma/at_hdmac.c | 79 ++++++++++++++++++++++++++++++++++++++++++++++++++
> 1 file changed, 79 insertions(+)
>
> diff --git a/drivers/dma/at_hdmac.c b/drivers/dma/at_hdmac.c
> index cad18f3660ae..4e55239c7a30 100644
> --- a/drivers/dma/at_hdmac.c
> +++ b/drivers/dma/at_hdmac.c
> @@ -986,6 +986,83 @@ err_free_buffer:
> return NULL;
> }
>
> +static struct dma_async_tx_descriptor *
> +atc_prep_dma_memset_sg(struct dma_chan *chan,
> + struct scatterlist *sgl,
> + unsigned int sg_len, int value,
> + unsigned long flags)
> +{
> + struct at_dma_chan *atchan = to_at_dma_chan(chan);
> + struct at_dma *atdma = to_at_dma(chan->device);
> + struct at_desc *desc = NULL, *first = NULL, *prev = NULL;
> + struct scatterlist *sg;
> + void __iomem *vaddr;
> + dma_addr_t paddr;
> + size_t total_len = 0;
> + int i;
> +
> + dev_vdbg(chan2dev(chan), "%s: v0x%x l0x%zx f0x%lx\n", __func__,
> + value, sg_len, flags);
> +
> + if (unlikely(!sgl || !sg_len)) {
> + dev_dbg(chan2dev(chan), "%s: scatterlist is empty!\n",
> + __func__);
> + return NULL;
> + }
> +
> + vaddr = dma_pool_alloc(atdma->memset_pool, GFP_ATOMIC, &paddr);
> + if (!vaddr) {
> + dev_err(chan2dev(chan), "%s: couldn't allocate buffer\n",
> + __func__);
> + return NULL;
> + }
> + *(u32*)vaddr = value;
> +
> + for_each_sg(sgl, sg, sg_len, i) {
> + dma_addr_t dest = sg_dma_address(sg);
> + size_t len = sg_dma_len(sg);
> +
> + dev_vdbg(chan2dev(chan), "%s: d0x%08x, l0x%zx\n",
> + __func__, dest, len);
> +
> + if (!is_dma_fill_aligned(chan->device, dest, 0, len)) {
> + dev_err(chan2dev(chan), "%s: buffer is not aligned\n",
> + __func__);
> + goto err_put_desc;
> + }
> +
> + desc = atc_create_memset_desc(chan, paddr, dest, len);
> + if (!desc)
> + goto err_put_desc;
> +
> + atc_desc_chain(&first, &prev, desc);
> +
> + total_len += len;
> + }
> +
> + /*
> + * Only set the buffer pointers on the last descriptor to
> + * avoid free'ing while we have our transfer still going
> + */
> + desc->memset_paddr = paddr;
> + desc->memset_vaddr = vaddr;
> + desc->memset_buffer = true;
> +
> + first->txd.cookie = -EBUSY;
> + first->total_len = total_len;
> +
> + /* set end-of-link on the descriptor */
> + set_desc_eol(desc);
> +
> + first->txd.flags = flags;
> +
> + return &first->txd;
> +
> +err_put_desc:
> + atc_desc_put(atchan, first);
> + return NULL;
> +}
> +
> /**
> * atc_prep_slave_sg - prepare descriptors for a DMA_SLAVE transaction
> * @chan: DMA channel
> @@ -1868,6 +1945,7 @@ static int __init at_dma_probe(struct platform_device *pdev)
> dma_cap_set(DMA_INTERLEAVE, at91sam9g45_config.cap_mask);
> dma_cap_set(DMA_MEMCPY, at91sam9g45_config.cap_mask);
> dma_cap_set(DMA_MEMSET, at91sam9g45_config.cap_mask);
> + dma_cap_set(DMA_MEMSET_SG, at91sam9g45_config.cap_mask);
> dma_cap_set(DMA_PRIVATE, at91sam9g45_config.cap_mask);
> dma_cap_set(DMA_SLAVE, at91sam9g45_config.cap_mask);
> dma_cap_set(DMA_SG, at91sam9g45_config.cap_mask);
> @@ -1989,6 +2067,7 @@ static int __init at_dma_probe(struct platform_device *pdev)
>
> if (dma_has_cap(DMA_MEMSET, atdma->dma_common.cap_mask)) {
> atdma->dma_common.device_prep_dma_memset = atc_prep_dma_memset;
> + atdma->dma_common.device_prep_dma_memset_sg = atc_prep_dma_memset_sg;
> atdma->dma_common.fill_align = DMAENGINE_ALIGN_4_BYTES;
> }
>
>
--
Nicolas Ferre
WARNING: multiple messages have this Message-ID (diff)
From: Nicolas Ferre <nicolas.ferre@atmel.com>
To: Maxime Ripard <maxime.ripard@free-electrons.com>,
Vinod Koul <vinod.koul@intel.com>,
Alexandre Belloni <alexandre.belloni@free-electrons.com>
Cc: <dmaengine@vger.kernel.org>,
Ludovic Desroches <ludovic.desroches@atmel.com>,
<linux-arm-kernel@lists.infradead.org>,
<linux-kernel@vger.kernel.org>
Subject: Re: [PATCH v2 2/2] dmaengine: hdmac: Add scatter-gathered memset support
Date: Thu, 22 Oct 2015 12:18:51 +0200 [thread overview]
Message-ID: <5628B80B.8040105@atmel.com> (raw)
In-Reply-To: <1445506860-22619-3-git-send-email-maxime.ripard@free-electrons.com>
Le 22/10/2015 11:41, Maxime Ripard a écrit :
> Just like memset support, the HDMAC might be used to do a memset over a
> discontiguous memory area.
>
> In such a case, we'll just build up a chain of memset descriptors over the
> contiguous chunks of memory to set, in order to allow such a support.
>
> Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
Yep, seems good!
Acked-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Thanks Maxime, bye.
> ---
> drivers/dma/at_hdmac.c | 79 ++++++++++++++++++++++++++++++++++++++++++++++++++
> 1 file changed, 79 insertions(+)
>
> diff --git a/drivers/dma/at_hdmac.c b/drivers/dma/at_hdmac.c
> index cad18f3660ae..4e55239c7a30 100644
> --- a/drivers/dma/at_hdmac.c
> +++ b/drivers/dma/at_hdmac.c
> @@ -986,6 +986,83 @@ err_free_buffer:
> return NULL;
> }
>
> +static struct dma_async_tx_descriptor *
> +atc_prep_dma_memset_sg(struct dma_chan *chan,
> + struct scatterlist *sgl,
> + unsigned int sg_len, int value,
> + unsigned long flags)
> +{
> + struct at_dma_chan *atchan = to_at_dma_chan(chan);
> + struct at_dma *atdma = to_at_dma(chan->device);
> + struct at_desc *desc = NULL, *first = NULL, *prev = NULL;
> + struct scatterlist *sg;
> + void __iomem *vaddr;
> + dma_addr_t paddr;
> + size_t total_len = 0;
> + int i;
> +
> + dev_vdbg(chan2dev(chan), "%s: v0x%x l0x%zx f0x%lx\n", __func__,
> + value, sg_len, flags);
> +
> + if (unlikely(!sgl || !sg_len)) {
> + dev_dbg(chan2dev(chan), "%s: scatterlist is empty!\n",
> + __func__);
> + return NULL;
> + }
> +
> + vaddr = dma_pool_alloc(atdma->memset_pool, GFP_ATOMIC, &paddr);
> + if (!vaddr) {
> + dev_err(chan2dev(chan), "%s: couldn't allocate buffer\n",
> + __func__);
> + return NULL;
> + }
> + *(u32*)vaddr = value;
> +
> + for_each_sg(sgl, sg, sg_len, i) {
> + dma_addr_t dest = sg_dma_address(sg);
> + size_t len = sg_dma_len(sg);
> +
> + dev_vdbg(chan2dev(chan), "%s: d0x%08x, l0x%zx\n",
> + __func__, dest, len);
> +
> + if (!is_dma_fill_aligned(chan->device, dest, 0, len)) {
> + dev_err(chan2dev(chan), "%s: buffer is not aligned\n",
> + __func__);
> + goto err_put_desc;
> + }
> +
> + desc = atc_create_memset_desc(chan, paddr, dest, len);
> + if (!desc)
> + goto err_put_desc;
> +
> + atc_desc_chain(&first, &prev, desc);
> +
> + total_len += len;
> + }
> +
> + /*
> + * Only set the buffer pointers on the last descriptor to
> + * avoid free'ing while we have our transfer still going
> + */
> + desc->memset_paddr = paddr;
> + desc->memset_vaddr = vaddr;
> + desc->memset_buffer = true;
> +
> + first->txd.cookie = -EBUSY;
> + first->total_len = total_len;
> +
> + /* set end-of-link on the descriptor */
> + set_desc_eol(desc);
> +
> + first->txd.flags = flags;
> +
> + return &first->txd;
> +
> +err_put_desc:
> + atc_desc_put(atchan, first);
> + return NULL;
> +}
> +
> /**
> * atc_prep_slave_sg - prepare descriptors for a DMA_SLAVE transaction
> * @chan: DMA channel
> @@ -1868,6 +1945,7 @@ static int __init at_dma_probe(struct platform_device *pdev)
> dma_cap_set(DMA_INTERLEAVE, at91sam9g45_config.cap_mask);
> dma_cap_set(DMA_MEMCPY, at91sam9g45_config.cap_mask);
> dma_cap_set(DMA_MEMSET, at91sam9g45_config.cap_mask);
> + dma_cap_set(DMA_MEMSET_SG, at91sam9g45_config.cap_mask);
> dma_cap_set(DMA_PRIVATE, at91sam9g45_config.cap_mask);
> dma_cap_set(DMA_SLAVE, at91sam9g45_config.cap_mask);
> dma_cap_set(DMA_SG, at91sam9g45_config.cap_mask);
> @@ -1989,6 +2067,7 @@ static int __init at_dma_probe(struct platform_device *pdev)
>
> if (dma_has_cap(DMA_MEMSET, atdma->dma_common.cap_mask)) {
> atdma->dma_common.device_prep_dma_memset = atc_prep_dma_memset;
> + atdma->dma_common.device_prep_dma_memset_sg = atc_prep_dma_memset_sg;
> atdma->dma_common.fill_align = DMAENGINE_ALIGN_4_BYTES;
> }
>
>
--
Nicolas Ferre
next prev parent reply other threads:[~2015-10-22 10:18 UTC|newest]
Thread overview: 12+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-10-22 9:40 [PATCH v2 0/2] dmaengine: hdmac: Add scatter-gathered memset support Maxime Ripard
2015-10-22 9:40 ` Maxime Ripard
2015-10-22 9:40 ` [PATCH v2 1/2] dmaengine: hdmac: factorise memset descriptor allocation Maxime Ripard
2015-10-22 9:40 ` Maxime Ripard
2015-10-22 10:13 ` Nicolas Ferre
2015-10-22 10:13 ` Nicolas Ferre
2015-10-22 9:41 ` [PATCH v2 2/2] dmaengine: hdmac: Add scatter-gathered memset support Maxime Ripard
2015-10-22 9:41 ` Maxime Ripard
2015-10-22 10:18 ` Nicolas Ferre [this message]
2015-10-22 10:18 ` Nicolas Ferre
2015-10-29 1:42 ` [PATCH v2 0/2] " Vinod Koul
2015-10-29 1:42 ` Vinod Koul
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=5628B80B.8040105@atmel.com \
--to=nicolas.ferre@atmel.com \
--cc=linux-arm-kernel@lists.infradead.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 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.