Op 04-03-2007 om 08:33 schreef Gordon Farquharson:
> --- flash-kernel.orig 2007-03-03 11:42:45.000000000 -0700
> +++ flash-kernel      2007-03-03 13:30:35.000000000 -0700
> @@ -131,10 +131,9 @@
>               size=$(grep "Ramdisk" /proc/mtd | cut -d " " -f 2)
>               size=$(printf "%d" 0x$size)
>               isize=$(wc -c $ifile | awk '{print $1}')
> -             cat $ifile > $tmp

unconditional overwrite of $tmp with $ifile is removed

>               pad=$(expr $size - $isize - 16)

caculate the remainder

>               if [ "$pad" -gt 0 ]; then

when there are padding bytes and only when there are padding bytes

> -                     dd if=/dev/zero bs=$pad count=1 2>/dev/null >> $tmp

the old append padding bytes instruction being removed

> +                     dd if=$ifile of=$tmp ibs=4M conv=sync 2>/dev/null

overwrite $tmp with the first four megabyte of $ifile

>               fi
>               (
>                       sercomm_header $isize



My gut feeling[1] says it should be

--- flash-kernel.orig   2007-03-03 11:42:45.000000000 -0700
+++ flash-kernel        2007-03-03 13:30:35.000000000 -0700
@@ -131,10 +131,6 @@
                size=$(grep "Ramdisk" /proc/mtd | cut -d " " -f 2)
                size=$(printf "%d" 0x$size)
                isize=$(wc -c $ifile | awk '{print $1}')
-               cat $ifile > $tmp
-               pad=$(expr $size - $isize - 16)
-               if [ "$pad" -gt 0 ]; then
-                       dd if=/dev/zero bs=$pad count=1 2>/dev/null >> $tmp
-               fi
+               dd if=$ifile of=$tmp ibs=4M conv=sync 2>/dev/null
                (
                        sercomm_header $isize

So $tmp is unconditional overwritten with $ifile

Actual is $tmp unconditional overwritten with the first 4Mb of $ifile,

but I have no clue to what the 4M corresponds.

Something else:
"isize" is not used in the proposed
+               dd if=$ifile of=$tmp ibs=4M conv=sync 2>/dev/null
not calculating $isize will also speed up flashing the kernel.


Cheers
Geert Stappers
[1] the patch is not tested.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Reply via email to