point()/unpoint() questions + small cfi_cmdset_0001.c patch
Joakim Tjernlund
Joakim.Tjernlund at lumentis.se
Mon Jun 17 05:24:33 EDT 2002
Hi all
I am looking into using point()/unpoint() in scan.c instead of mtd->read(). A few questions:
1) There is no point()/unpoint() function pointers in struct map_info, so I wounder where one should
add the point()/unpoint() functions?
2) It is not clear to me how the semantics for point()/unpoint() works, is the following sequence legal:
point()
...
read()
...
point()
....
point()
....
unpoint()
...
write()
Also here is a small patch which fixes 2 bugs:
1) Flash always reports these in litte endian. This corrects things on a BE CPU.
2) Was doing if ( (!*p) & 2), should be if (! (*p &2))
Index: cfi_cmdset_0001.c
===================================================================
RCS file: /home/cvs/mtd/drivers/mtd/chips/cfi_cmdset_0001.c,v
retrieving revision 1.98
diff -u -r1.98 cfi_cmdset_0001.c
--- cfi_cmdset_0001.c 16 Jun 2002 17:14:06 -0000 1.98
+++ cfi_cmdset_0001.c 17 Jun 2002 09:17:05 -0000
@@ -156,9 +156,9 @@
}
/* Do some byteswapping if necessary */
- extp->FeatureSupport = cfi32_to_cpu(extp->FeatureSupport);
- extp->BlkStatusRegMask = cfi32_to_cpu(extp->BlkStatusRegMask);
- extp->ProtRegAddr = cfi32_to_cpu(extp->ProtRegAddr);
+ extp->FeatureSupport = le32_to_cpu(extp->FeatureSupport);
+ extp->BlkStatusRegMask = le16_to_cpu(extp->BlkStatusRegMask);
+ extp->ProtRegAddr = le16_to_cpu(extp->ProtRegAddr);
#ifdef DEBUG_CFI_FEATURES
/* Tell the user about it in lots of lovely detail */
@@ -300,7 +300,7 @@
*/
switch (chip->state) {
case FL_ERASING:
- if (!((struct cfi_pri_intelext *)cfi->cmdset_priv)->FeatureSupport & 2)
+ if (!(((struct cfi_pri_intelext *)cfi->cmdset_priv)->FeatureSupport & 2))
goto sleep; /* We don't support erase suspend */
cfi_write (map, CMD(0xb0), cmd_addr);
More information about the linux-mtd
mailing list