Patchwork Help: unable to upgrade bios

login
register
about
Submitter David Hendricks
Date 2012-01-10 22:01:50
Message ID <CAKOoyUdpsDiA47ghrmFA7kt2xTmkxCm2F8AE84+TaYaN7GQX9Q@mail.gmail.com>
Download mbox | patch
Permalink /patch/3501/
State Not for merge
Headers show

Comments

David Hendricks - 2012-01-10 22:01:50
On Mon, Jan 9, 2012 at 3:20 AM, Raghuramchary Jallipalli <
Raghuramchary.Jallipalli@lntinfotech.com> wrote:

>  Thanks David.****
>
> As you said, the reason for not programming is due to wrong op.atomic
> configuration. Is there any way to set these properly?
>

It should be set automatically.

Carl-Daniel clarified on IRC that the "atomic" cycle flag will always be 0
at the stage they are printed. Sorry for the confusion!

I recommend adding a debug print to ich9_run_opcode() to see what the
"atomic" flag ends up being:

                /* Select second preop. */

Then run "flashrom -V -p internal:this_is_not_a_laptop -r foo.bin" and make
sure that the sector erase opcode (0x20) and byte program opcode (0x02)
show either "1" or "2" for op.atomic.

This will help us to prove whether or not this is the problem you are
encountering.

****
>
> Yes jumper selects the ROM at a time.
>

Okay. I do not think Flashrom is confused about the flash part, then.
Raghuramchary Jallipalli - 2012-02-02 08:42:26
Hi David,
Sorry for the late reply.
I have done the changes as suggested and don’t see either 1 or 2 for op.atomic after issuing the “flashrom –V –p internal:this_is_not_a_laptop –r foo.bin” command. Please see attached log.

Will this now confirm that the flashrom utility cannot be used for BIOS upgrade on my board?

NOTE: The log is collected for AT25DF641A(8MB) flash as opposed to AT25DF321A(4MB) flash before.

Thanks for the help,
Raghu

From: David Hendricks [mailto:dhendrix@google.com]
Sent: Wednesday, January 11, 2012 3:32 AM
To: Raghuramchary Jallipalli
Cc: flashrom@flashrom.org; V Balaji; Vishnuvardhan Reddy
Subject: Re: [flashrom] Help: unable to upgrade bios

On Mon, Jan 9, 2012 at 3:20 AM, Raghuramchary Jallipalli <Raghuramchary.Jallipalli@lntinfotech.com<mailto:Raghuramchary.Jallipalli@lntinfotech.com>> wrote:
Thanks David.
As you said, the reason for not programming is due to wrong op.atomic configuration. Is there any way to set these properly?

It should be set automatically.

Carl-Daniel clarified on IRC that the "atomic" cycle flag will always be 0 at the stage they are printed. Sorry for the confusion!

I recommend adding a debug print to ich9_run_opcode() to see what the "atomic" flag ends up being:


Then run "flashrom -V -p internal:this_is_not_a_laptop -r foo.bin" and make sure that the sector erase opcode (0x20) and byte program opcode (0x02) show either "1" or "2" for op.atomic.

This will help us to prove whether or not this is the problem you are encountering.

Yes jumper selects the ROM at a time.

Okay. I do not think Flashrom is confused about the flash part, then.

--
David Hendricks (dhendrix)
Systems Software Engineer, Google Inc.

Patch

Index: ichspi.c
===================================================================
--- ichspi.c    (revision 1483)
+++ ichspi.c    (working copy)
@@ -908,6 +908,7 @@ 
           This means the timeout must be sufficient for chip erase
           of slow high-capacity chips.
         */
+       msg_pdbg("opcode: 0x%02x, spi_type: %u, atomic: %u\n", op.opcode,
op.spi_type, op.atomic);
        switch (op.atomic) {
        case 2: