提交 5aa2110f 编写于 作者: G Guennadi Liakhovetski 提交者: Mauro Carvalho Chehab

V4L/DVB (7671): pxa-camera: fix DMA sg-list coalescing for more than 2 buffers

Currently the pxa-camera driver has a bug, visible when the user requests
more than 2 video buffers. When the third buffer is queued, it is not
appended to the DMA-descriptor list of the second buffer, but is again
appended to the first buffer. Fix it.
Signed-off-by: NGuennadi Liakhovetski <g.liakhovetski@gmx.de>
Signed-off-by: NMauro Carvalho Chehab <mchehab@infradead.org>
上级 e7c50688
......@@ -102,11 +102,6 @@ struct pxa_buffer {
enum pxa_camera_active_dma active_dma;
};
struct pxa_framebuffer_queue {
dma_addr_t sg_last_dma;
struct pxa_dma_desc *sg_last_cpu;
};
struct pxa_camera_dev {
struct device *dev;
/* PXA27x is only supposed to handle one camera on its Quick Capture
......@@ -131,6 +126,7 @@ struct pxa_camera_dev {
spinlock_t lock;
struct pxa_buffer *active;
struct pxa_dma_desc *sg_tail[3];
};
static const char *pxa_cam_driver_description = "PXA_Camera";
......@@ -144,11 +140,14 @@ static int pxa_videobuf_setup(struct videobuf_queue *vq, unsigned int *count,
unsigned int *size)
{
struct soc_camera_device *icd = vq->priv_data;
struct soc_camera_host *ici =
to_soc_camera_host(icd->dev.parent);
struct pxa_camera_dev *pcdev = ici->priv;
dev_dbg(&icd->dev, "count=%d, size=%d\n", *count, *size);
/* planar capture requires Y, U and V buffers to be page aligned */
if (icd->current_fmt->fourcc == V4L2_PIX_FMT_YUV422P) {
if (pcdev->channels == 3) {
*size = PAGE_ALIGN(icd->width * icd->height); /* Y pages */
*size += PAGE_ALIGN(icd->width * icd->height / 2); /* U pages */
*size += PAGE_ALIGN(icd->width * icd->height / 2); /* V pages */
......@@ -296,7 +295,7 @@ static int pxa_videobuf_prepare(struct videobuf_queue *vq,
if (ret)
goto fail;
if (buf->fmt->fourcc == V4L2_PIX_FMT_YUV422P) {
if (pcdev->channels == 3) {
/* FIXME the calculations should be more precise */
sglen_y = dma->sglen / 2;
sglen_u = sglen_v = dma->sglen / 4 + 1;
......@@ -318,7 +317,7 @@ static int pxa_videobuf_prepare(struct videobuf_queue *vq,
goto fail;
}
if (buf->fmt->fourcc == V4L2_PIX_FMT_YUV422P) {
if (pcdev->channels == 3) {
/* init DMA for U channel */
ret = pxa_init_dma_channel(pcdev, buf, dma, 1, sglen_u,
sglen_y, 0x30, size_u);
......@@ -343,7 +342,7 @@ static int pxa_videobuf_prepare(struct videobuf_queue *vq,
buf->inwork = 0;
buf->active_dma = DMA_Y;
if (buf->fmt->fourcc == V4L2_PIX_FMT_YUV422P)
if (pcdev->channels == 3)
buf->active_dma |= DMA_U | DMA_V;
return 0;
......@@ -371,6 +370,7 @@ static void pxa_videobuf_queue(struct videobuf_queue *vq,
struct pxa_buffer *buf = container_of(vb, struct pxa_buffer, vb);
struct pxa_buffer *active;
unsigned long flags;
int i;
dev_dbg(&icd->dev, "%s (vb=0x%p) 0x%08lx %d\n", __func__,
vb, vb->baddr, vb->bsize);
......@@ -383,15 +383,11 @@ static void pxa_videobuf_queue(struct videobuf_queue *vq,
if (!active) {
CIFR |= CIFR_RESET_F;
DDADR(pcdev->dma_chans[0]) = buf->dmas[0].sg_dma;
DCSR(pcdev->dma_chans[0]) = DCSR_RUN;
if (buf->fmt->fourcc == V4L2_PIX_FMT_YUV422P) {
DDADR(pcdev->dma_chans[1]) = buf->dmas[1].sg_dma;
DCSR(pcdev->dma_chans[1]) = DCSR_RUN;
DDADR(pcdev->dma_chans[2]) = buf->dmas[2].sg_dma;
DCSR(pcdev->dma_chans[2]) = DCSR_RUN;
for (i = 0; i < pcdev->channels; i++) {
DDADR(pcdev->dma_chans[i]) = buf->dmas[i].sg_dma;
DCSR(pcdev->dma_chans[i]) = DCSR_RUN;
pcdev->sg_tail[i] = buf->dmas[i].sg_cpu + buf->dmas[i].sglen - 1;
}
pcdev->active = buf;
......@@ -400,7 +396,6 @@ static void pxa_videobuf_queue(struct videobuf_queue *vq,
struct pxa_cam_dma *buf_dma;
struct pxa_cam_dma *act_dma;
int nents;
int i;
for (i = 0; i < pcdev->channels; i++) {
buf_dma = &buf->dmas[i];
......@@ -412,8 +407,8 @@ static void pxa_videobuf_queue(struct videobuf_queue *vq,
/* Add the descriptors we just initialized to
the currently running chain */
act_dma->sg_cpu[act_dma->sglen - 1].ddadr =
buf_dma->sg_dma;
pcdev->sg_tail[i]->ddadr = buf_dma->sg_dma;
pcdev->sg_tail[i] = buf_dma->sg_cpu + buf_dma->sglen - 1;
/* Setup a dummy descriptor with the DMA engines current
* state
......
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册