Em 30-12-2010 13:21, Hans Verkuil escreveu: > On Thursday, December 30, 2010 16:00:41 Mauro Carvalho Chehab wrote: >> Fix two gcc warnings: >> >> drivers/media/video/ivtv/ivtv-i2c.c:170: warning: cast from pointer to integer of different size >> drivers/media/video/ivtv/ivtv-i2c.c:171: warning: cast from pointer to integer of different size >> $ gcc --version >> gcc (GCC) 4.1.2 20080704 (Red Hat 4.1.2-48) >> >> They seem bogus, but, as the original code also has problems with >> LE/BE, just change its implementation to be clear. > > Definitely not bogus: Yes, it is not bogus. > > unsigned char keybuf[4]; > > .. > > *ir_key = (u32) keybuf; > > Here keybuf == &keybuf[0]. So you put the address of keybuf in *ir_key. Which is > indeed of a different size in the case of a 64-bit architecture. > > What you probably meant to do is: > > *ir_key = *(u32 *)keybuf; Yes, but this would lead into BE/LE issues. > > Note that the code in your patch assumes that keybuf is in big-endian order. I assume > that's what it should be? Based on the way lirc_i2c was printing the data, this seems to be the most coherent way. We'll only know for sure after testing the remote with the hardware. > > Regards, > > Hans > >> Signed-off-by: Mauro Carvalho Chehab <mchehab@xxxxxxxxxx> >> >> diff --git a/drivers/media/video/ivtv/ivtv-i2c.c b/drivers/media/video/ivtv/ivtv-i2c.c >> index 2bed430..e103b8f 100644 >> --- a/drivers/media/video/ivtv/ivtv-i2c.c >> +++ b/drivers/media/video/ivtv/ivtv-i2c.c >> @@ -167,8 +167,8 @@ static int get_key_adaptec(struct IR_i2c *ir, u32 *ir_key, u32 *ir_raw) >> keybuf[2] &= 0x7f; >> keybuf[3] |= 0x80; >> >> - *ir_key = (u32) keybuf; >> - *ir_raw = (u32) keybuf; >> + *ir_key = keybuf[3] | keybuf[2] << 8 | keybuf[1] << 16 |keybuf[0] << 24; >> + *ir_raw = *ir_key; >> >> return 1; >> } >> > -- To unsubscribe from this list: send the line "unsubscribe linux-media" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
