Re: [flashrom] [PATCH] Move implicit erase out of chip drivers, clean up

[Uwe Hermann]

On Sat, Jul 17, 2010 at 12:18:49AM +0200, Carl-Daniel Hailfinger wrote:
> New version, updated to apply against current svn.
>
> flashrom had an implicit erase-on-write for most flash chip and

^
                                                             s
> Compile tested and proofread, but that's it. The patch may cause
> flashrom to eat your dog or it may cause your dog to eat burned flash chips.
> Please test on real hardware. Thanks!
>
> I have a cleanup patch for the write functions which will apply on top
> of this, but I wanted to keep this patch readable and self-contained.
>
> Signed-off-by: Carl-Daniel Hailfinger

Acked-by: Uwe Hermann

Tested on ASUS A8NE-FM with "SST SST49LF004A/B" (512 KB, FWH) and on a
Dr. Kaiser PCI card with a "Macronix MX29F001B" (128 KB, Parallel).

Tested all operations including erase/write multiple times for good measure.
> Index: flashrom-no_implicit_erase_inside_write/flashchips.c
>
> --- flashrom-no_implicit_erase_inside_write/flashchips.c	(Revision 1083)
> +++ flashrom-no_implicit_erase_inside_write/flashchips.c	(Arbeitskopie)
> @@ -453,7 +453,7 @@
>    	.model_id	= AT_25DF321,
>    	.total_size	= 4096,
>    	.page_size	= 256,
> .tested		= TEST_OK_PRW,
> .tested		= TEST_OK_PR,
> .probe		= probe_spi_rdid,
> .probe_timing	= TIMING_ZERO,
> .block_erasers	=

Didn't check whether any new chips were added since this patch which
would also need their W bit removed. Feel free to merge that into the
patch and re-use my ack.
> Index: flashrom-no_implicit_erase_inside_write/82802ab.c
>
> --- flashrom-no_implicit_erase_inside_write/82802ab.c	(Revision 1083)
> +++ flashrom-no_implicit_erase_inside_write/82802ab.c	(Arbeitskopie)
> @@ -144,23 +144,6 @@
>    return 0;
>  }
>
> -int erase_82802ab(struct flashchip *flash)
> -{
> int i;
> unsigned int total_size = flash->total_size * 1024;
> msg_cspew("total_size is %d; flash->page_size is %d\n",
>       total_size, flash->page_size);
> for (i = 0; i < total_size; i += flash->page_size)
> if (erase_block_82802ab(flash, i, flash->page_size)) {
> 	msg_cerr("ERASE FAILED!\n");
> 	return -1;
> }
> msg_cinfo("DONE ERASE\n");
> return 0;
>
> -}

No longer needed? Not used currently anywhere, it seems.

Uwe.