From: "Edgar E. Iglesias" <edgar.iglesias@xilinx.com>
To: Sai Pavan Boddu <sai.pavan.boddu@xilinx.com>
Cc: qemu-devel@nongnu.org,
"Alistair Francis" <alistair@alistair23.me>,
"Peter Crosthwaite" <crosthwaite.peter@gmail.com>,
"Peter Maydell" <peter.maydell@linaro.org>,
"Peter A. G. Crosthwaite" <peter.crosthwaite@petalogix.com>,
"Francisco Iglesias" <frasse.iglesias@gmail.com>,
"Philippe Mathieu-Daudé" <f4bug@amsat.org>
Subject: Re: [Qemu-devel] [PATCH v2] xilinx_spips: Make dma transactions as per dma_burst_size
Date: Wed, 20 Jun 2018 10:07:18 +0200 [thread overview]
Message-ID: <20180620080718.5dak7rys37yluigo@toto> (raw)
In-Reply-To: <1529440203-29407-1-git-send-email-saipava@xilinx.com>
On Wed, Jun 20, 2018 at 02:00:00AM +0530, Sai Pavan Boddu wrote:
> Qspi dma has a burst length of 64 bytes, So limit transaction length to
> 64 max.
>
> Signed-off-by: Sai Pavan Boddu <saipava@xilinx.com>
> ---
> hw/ssi/xilinx_spips.c | 20 +++++++++++++++++---
> include/hw/ssi/xilinx_spips.h | 5 ++++-
> 2 files changed, 21 insertions(+), 4 deletions(-)
>
> diff --git a/hw/ssi/xilinx_spips.c b/hw/ssi/xilinx_spips.c
> index 03f5fae..0ea57d1 100644
> --- a/hw/ssi/xilinx_spips.c
> +++ b/hw/ssi/xilinx_spips.c
> @@ -851,12 +851,17 @@ static void xlnx_zynqmp_qspips_notify(void *opaque)
> {
> size_t ret;
> uint32_t num;
> - const void *rxd = pop_buf(recv_fifo, 4, &num);
> + const void *rxd;
> + int len;
> +
> + len = recv_fifo->num >= rq->dma_burst_size ? rq->dma_burst_size :
> + recv_fifo->num;
> + rxd = pop_buf(recv_fifo, len, &num);
>
> memcpy(rq->dma_buf, rxd, num);
>
> - ret = stream_push(rq->dma, rq->dma_buf, 4);
> - assert(ret == 4);
> + ret = stream_push(rq->dma, rq->dma_buf, num);
> + assert(ret == num);
> xlnx_zynqmp_qspips_check_flush(rq);
> }
> }
> @@ -1337,6 +1342,9 @@ static void xlnx_zynqmp_qspips_realize(DeviceState *dev, Error **errp)
> fifo8_create(&s->rx_fifo_g, xsc->rx_fifo_size);
> fifo8_create(&s->tx_fifo_g, xsc->tx_fifo_size);
> fifo32_create(&s->fifo_g, 32);
> + if (s->dma_burst_size > QSPI_DMA_MAX_BURST_SIZE) {
> + s->dma_burst_size = QSPI_DMA_MAX_BURST_SIZE;
> + }
I think it's better to let the user know that the chosen value was out of
range, look at how num-busses is handled at the begining of this function.
> }
>
> static void xlnx_zynqmp_qspips_init(Object *obj)
> @@ -1411,6 +1419,11 @@ static const VMStateDescription vmstate_xlnx_zynqmp_qspips = {
> }
> };
>
> +static Property xilinx_zynqmp_qspips_properties[] = {
> + DEFINE_PROP_UINT32("dma-burst-size", XlnxZynqMPQSPIPS, dma_burst_size, 64),
> + DEFINE_PROP_END_OF_LIST(),
> +};
> +
> static Property xilinx_qspips_properties[] = {
> /* We had to turn this off for 2.10 as it is not compatible with migration.
> * It can be enabled but will prevent the device to be migrated.
> @@ -1463,6 +1476,7 @@ static void xlnx_zynqmp_qspips_class_init(ObjectClass *klass, void * data)
> dc->realize = xlnx_zynqmp_qspips_realize;
> dc->reset = xlnx_zynqmp_qspips_reset;
> dc->vmsd = &vmstate_xlnx_zynqmp_qspips;
> + dc->props = xilinx_zynqmp_qspips_properties;
> xsc->reg_ops = &xlnx_zynqmp_qspips_ops;
> xsc->rx_fifo_size = RXFF_A_Q;
> xsc->tx_fifo_size = TXFF_A_Q;
> diff --git a/include/hw/ssi/xilinx_spips.h b/include/hw/ssi/xilinx_spips.h
> index d398a4e..bc5596a 100644
> --- a/include/hw/ssi/xilinx_spips.h
> +++ b/include/hw/ssi/xilinx_spips.h
> @@ -37,6 +37,8 @@ typedef struct XilinxSPIPS XilinxSPIPS;
> /* Bite off 4k chunks at a time */
> #define LQSPI_CACHE_SIZE 1024
>
> +#define QSPI_DMA_MAX_BURST_SIZE 2048
> +
> typedef enum {
> READ = 0x3, READ_4 = 0x13,
> FAST_READ = 0xb, FAST_READ_4 = 0x0c,
> @@ -95,7 +97,8 @@ typedef struct {
> XilinxQSPIPS parent_obj;
>
> StreamSlave *dma;
> - uint8_t dma_buf[4];
> + uint8_t dma_buf[QSPI_DMA_MAX_BURST_SIZE];
> + uint32_t dma_burst_size;
> int gqspi_irqline;
>
> uint32_t regs[XLNX_ZYNQMP_SPIPS_R_MAX];
> --
> 2.7.4
>
prev parent reply other threads:[~2018-06-20 8:07 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-06-19 20:30 [Qemu-devel] [PATCH v2] xilinx_spips: Make dma transactions as per dma_burst_size Sai Pavan Boddu
2018-06-20 8:07 ` Edgar E. Iglesias [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=20180620080718.5dak7rys37yluigo@toto \
--to=edgar.iglesias@xilinx.com \
--cc=alistair@alistair23.me \
--cc=crosthwaite.peter@gmail.com \
--cc=f4bug@amsat.org \
--cc=frasse.iglesias@gmail.com \
--cc=peter.crosthwaite@petalogix.com \
--cc=peter.maydell@linaro.org \
--cc=qemu-devel@nongnu.org \
--cc=sai.pavan.boddu@xilinx.com \
/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).