* [patch 2/2] [media] s2255drv: fix some endian bugs
@ 2012-02-17 6:44 Dan Carpenter
2012-02-23 0:37 ` dean anderson
0 siblings, 1 reply; 2+ messages in thread
From: Dan Carpenter @ 2012-02-17 6:44 UTC (permalink / raw)
To: Mauro Carvalho Chehab
Cc: Dean Anderson, Hans Verkuil, Pete Eberlein, linux-media,
kernel-janitors
I don't have this hardware and I don't know the subsystem very well. So
please review this patch carefully. The original code definitely looks
buggy though.
Sparse complains about some endian bugs where little endian bugs are
treated as cpu endian.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
---
diff --git a/drivers/media/video/s2255drv.c b/drivers/media/video/s2255drv.c
index 3505242..4894cbb 100644
--- a/drivers/media/video/s2255drv.c
+++ b/drivers/media/video/s2255drv.c
@@ -134,7 +134,7 @@
/* usb config commands */
#define IN_DATA_TOKEN cpu_to_le32(0x2255c0de)
-#define CMD_2255 cpu_to_le32(0xc2255000)
+#define CMD_2255 0xc2255000
#define CMD_SET_MODE cpu_to_le32((CMD_2255 | 0x10))
#define CMD_START cpu_to_le32((CMD_2255 | 0x20))
#define CMD_STOP cpu_to_le32((CMD_2255 | 0x30))
@@ -2025,7 +2025,7 @@ static int save_frame(struct s2255_dev *dev, struct s2255_pipeinfo *pipe_info)
pdata[1]);
offset = jj + PREFIX_SIZE;
bframe = 1;
- cc = pdword[1];
+ cc = le32_to_cpu(pdword[1]);
if (cc >= MAX_CHANNELS) {
printk(KERN_ERR
"bad channel\n");
@@ -2034,22 +2034,22 @@ static int save_frame(struct s2255_dev *dev, struct s2255_pipeinfo *pipe_info)
/* reverse it */
dev->cc = G_chnmap[cc];
channel = &dev->channel[dev->cc];
- payload = pdword[3];
+ payload = le32_to_cpu(pdword[3]);
if (payload > channel->req_image_size) {
channel->bad_payload++;
/* discard the bad frame */
return -EINVAL;
}
channel->pkt_size = payload;
- channel->jpg_size = pdword[4];
+ channel->jpg_size = le32_to_cpu(pdword[4]);
break;
case S2255_MARKER_RESPONSE:
pdata += DEF_USB_BLOCK;
jj += DEF_USB_BLOCK;
- if (pdword[1] >= MAX_CHANNELS)
+ if (le32_to_cpu(pdword[1]) >= MAX_CHANNELS)
break;
- cc = G_chnmap[pdword[1]];
+ cc = G_chnmap[le32_to_cpu(pdword[1])];
if (cc >= MAX_CHANNELS)
break;
channel = &dev->channel[cc];
@@ -2072,11 +2072,11 @@ static int save_frame(struct s2255_dev *dev, struct s2255_pipeinfo *pipe_info)
wake_up(&dev->fw_data->wait_fw);
break;
case S2255_RESPONSE_STATUS:
- channel->vidstatus = pdword[3];
+ channel->vidstatus = le32_to_cpu(pdword[3]);
channel->vidstatus_ready = 1;
wake_up(&channel->wait_vidstatus);
dprintk(5, "got vidstatus %x chan %d\n",
- pdword[3], cc);
+ le32_to_cpu(pdword[3]), cc);
break;
default:
printk(KERN_INFO "s2255 unknown resp\n");
@@ -2603,10 +2603,11 @@ static int s2255_probe(struct usb_interface *interface,
__le32 *pRel;
pRel = (__le32 *) &dev->fw_data->fw->data[fw_size - 4];
printk(KERN_INFO "s2255 dsp fw version %x\n", *pRel);
- dev->dsp_fw_ver = *pRel;
- if (*pRel < S2255_CUR_DSP_FWVER)
+ dev->dsp_fw_ver = le32_to_cpu(*pRel);
+ if (dev->dsp_fw_ver < S2255_CUR_DSP_FWVER)
printk(KERN_INFO "s2255: f2255usb.bin out of date.\n");
- if (dev->pid = 0x2257 && *pRel < S2255_MIN_DSP_COLORFILTER)
+ if (dev->pid = 0x2257 &&
+ dev->dsp_fw_ver < S2255_MIN_DSP_COLORFILTER)
printk(KERN_WARNING "s2255: 2257 requires firmware %d"
" or above.\n", S2255_MIN_DSP_COLORFILTER);
}
^ permalink raw reply related [flat|nested] 2+ messages in thread* Re: [patch 2/2] [media] s2255drv: fix some endian bugs
2012-02-17 6:44 [patch 2/2] [media] s2255drv: fix some endian bugs Dan Carpenter
@ 2012-02-23 0:37 ` dean anderson
0 siblings, 0 replies; 2+ messages in thread
From: dean anderson @ 2012-02-23 0:37 UTC (permalink / raw)
To: Dan Carpenter
Cc: Mauro Carvalho Chehab, Dean Anderson, Hans Verkuil, Pete Eberlein,
linux-media, kernel-janitors
Hi Dan,
The original code has an issue on big endian hardware. The patch looks ok.
Signed-off-by: Dean Anderson <linux-dev@sensoray.com>
On 2/16/2012 10:44 PM, Dan Carpenter wrote:
> I don't have this hardware and I don't know the subsystem very well. So
> please review this patch carefully. The original code definitely looks
> buggy though.
>
> Sparse complains about some endian bugs where little endian bugs are
> treated as cpu endian.
>
> Signed-off-by: Dan Carpenter<dan.carpenter@oracle.com>
> ---
>
> diff --git a/drivers/media/video/s2255drv.c b/drivers/media/video/s2255drv.c
> index 3505242..4894cbb 100644
> --- a/drivers/media/video/s2255drv.c
> +++ b/drivers/media/video/s2255drv.c
> @@ -134,7 +134,7 @@
>
> /* usb config commands */
> #define IN_DATA_TOKEN cpu_to_le32(0x2255c0de)
> -#define CMD_2255 cpu_to_le32(0xc2255000)
> +#define CMD_2255 0xc2255000
> #define CMD_SET_MODE cpu_to_le32((CMD_2255 | 0x10))
> #define CMD_START cpu_to_le32((CMD_2255 | 0x20))
> #define CMD_STOP cpu_to_le32((CMD_2255 | 0x30))
> @@ -2025,7 +2025,7 @@ static int save_frame(struct s2255_dev *dev, struct s2255_pipeinfo *pipe_info)
> pdata[1]);
> offset = jj + PREFIX_SIZE;
> bframe = 1;
> - cc = pdword[1];
> + cc = le32_to_cpu(pdword[1]);
> if (cc>= MAX_CHANNELS) {
> printk(KERN_ERR
> "bad channel\n");
> @@ -2034,22 +2034,22 @@ static int save_frame(struct s2255_dev *dev, struct s2255_pipeinfo *pipe_info)
> /* reverse it */
> dev->cc = G_chnmap[cc];
> channel =&dev->channel[dev->cc];
> - payload = pdword[3];
> + payload = le32_to_cpu(pdword[3]);
> if (payload> channel->req_image_size) {
> channel->bad_payload++;
> /* discard the bad frame */
> return -EINVAL;
> }
> channel->pkt_size = payload;
> - channel->jpg_size = pdword[4];
> + channel->jpg_size = le32_to_cpu(pdword[4]);
> break;
> case S2255_MARKER_RESPONSE:
>
> pdata += DEF_USB_BLOCK;
> jj += DEF_USB_BLOCK;
> - if (pdword[1]>= MAX_CHANNELS)
> + if (le32_to_cpu(pdword[1])>= MAX_CHANNELS)
> break;
> - cc = G_chnmap[pdword[1]];
> + cc = G_chnmap[le32_to_cpu(pdword[1])];
> if (cc>= MAX_CHANNELS)
> break;
> channel =&dev->channel[cc];
> @@ -2072,11 +2072,11 @@ static int save_frame(struct s2255_dev *dev, struct s2255_pipeinfo *pipe_info)
> wake_up(&dev->fw_data->wait_fw);
> break;
> case S2255_RESPONSE_STATUS:
> - channel->vidstatus = pdword[3];
> + channel->vidstatus = le32_to_cpu(pdword[3]);
> channel->vidstatus_ready = 1;
> wake_up(&channel->wait_vidstatus);
> dprintk(5, "got vidstatus %x chan %d\n",
> - pdword[3], cc);
> + le32_to_cpu(pdword[3]), cc);
> break;
> default:
> printk(KERN_INFO "s2255 unknown resp\n");
> @@ -2603,10 +2603,11 @@ static int s2255_probe(struct usb_interface *interface,
> __le32 *pRel;
> pRel = (__le32 *)&dev->fw_data->fw->data[fw_size - 4];
> printk(KERN_INFO "s2255 dsp fw version %x\n", *pRel);
> - dev->dsp_fw_ver = *pRel;
> - if (*pRel< S2255_CUR_DSP_FWVER)
> + dev->dsp_fw_ver = le32_to_cpu(*pRel);
> + if (dev->dsp_fw_ver< S2255_CUR_DSP_FWVER)
> printk(KERN_INFO "s2255: f2255usb.bin out of date.\n");
> - if (dev->pid = 0x2257&& *pRel< S2255_MIN_DSP_COLORFILTER)
> + if (dev->pid = 0x2257&&
> + dev->dsp_fw_ver< S2255_MIN_DSP_COLORFILTER)
> printk(KERN_WARNING "s2255: 2257 requires firmware %d"
> " or above.\n", S2255_MIN_DSP_COLORFILTER);
> }
>
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2012-02-23 0:37 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2012-02-17 6:44 [patch 2/2] [media] s2255drv: fix some endian bugs Dan Carpenter
2012-02-23 0:37 ` dean anderson
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox