-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
Hello all,
I reworked the physmem mapping and also did a hack to allow RO mapping of first 1MB which is not allowed under DPMI ;) I simply copy the memory and return a pointer to a copied data. This is used for the coreboot tables and there is no need for RW.
The dmidecode could be also DOS recompiled but executing that if quite complicated. Therefore maybe own parser for DMI would be nice.
Attached patches are for the flashrom and pciutils.
To make it work try following:
1) get RPMs of the cross compiler from DJGPP site
djcross-binutils-2.19.1-10ap.i386.rpm djcross-gcc-tools-4.4.1-1ap.i686.rpm djcross-gcc-4.3.2-8ap.i686.rpm
Somehow install them on your system. I used alien on 32bit machine to make debs and I forced them to install on amd64 ;)
2) Download all stuff from: http://assembler.cz/flashrom
3) patch flashrom source and pciutils source
4) compile the libgetopt in its dir
5) compile the pciutils, use the README.DJGPP
6) flashrom patch expects that the ../libpci contains the stuff what is installed from pciutils and ../libgetopt contains compiled getopt.a compile it too ;)
I put there also the binary image and cwsdpmi zip which exe files are needed to run the flashrom.exe (I think only cwdpmi.exe)
Have fun and please test ;)
The dos.patch contains some flashrom NULL ptr fix and also Carl-Daniels changes for Makefile.
Rudolf
Hi Rudolf,
thanks for your patch.
On 14.03.2010 18:36, Rudolf Marek wrote:
I reworked the physmem mapping and also did a hack to allow RO mapping of first 1MB which is not allowed under DPMI ;) I simply copy the memory and return a pointer to a copied data. This is used for the coreboot tables and there is no need for RW.
Neat trick.
The dmidecode could be also DOS recompiled but executing that if quite complicated. Therefore maybe own parser for DMI would be nice.
Agreed.
Attached patches are for the flashrom and pciutils.
To make it work try following: [...]
I added a short version of your description to the flashrom README.
Have fun and please test ;)
The dos.patch contains some flashrom NULL ptr fix and also Carl-Daniels changes for Makefile.
I made some more changes to the Makefile. "make dos" is something I'd rather avoid because it hardcodes too many assumptions.
Index: physmap.c
--- physmap.c (revision 935) +++ physmap.c (working copy) @@ -4,6 +4,7 @@
- Copyright (C) 2009 Peter Stuge peter@stuge.se
- Copyright (C) 2009 coresystems GmbH
- Copyright (C) 2010 Carl-Daniel Hailfinger
- Copyright (C) 2010 Rudolf Marek r.marek@assembler.cz
- This program is free software; you can redistribute it and/or modify
- it under the terms of the GNU General Public License as published by
@@ -27,7 +28,77 @@ #include <errno.h> #include "flash.h"
-#ifdef __DARWIN__ +#ifdef __DJGPP__ [...] +#elif __DARWIN__
Can you change this to #elif defined(__DARWIN) to keep the logic as it was? I don't have a Mac OS X machine, so I can't test...
#include <DirectIO/darwinio.h>
#define MEM_DEV "DirectIO"
If you drop the board_enable.c hunk (and if my latest Makefile+README patch is OK for you), this is Acked-by: Carl-Daniel Hailfinger c-d.hailfinger.devel.2006@gmx.net
I can update the README once your pciutils patch has been merged.
Regards, Carl-Daniel
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
I added a short version of your description to the flashrom README.
Ok
I made some more changes to the Makefile. "make dos" is something I'd rather avoid because it hardcodes too many assumptions.
I would prefer to have there something like:
make djgpp-dos
because I will tend to forgot how to invoke the make from command line.
Index: physmap.c
--- physmap.c (revision 935) +++ physmap.c (working copy) @@ -4,6 +4,7 @@
- Copyright (C) 2009 Peter Stuge peter@stuge.se
- Copyright (C) 2009 coresystems GmbH
- Copyright (C) 2010 Carl-Daniel Hailfinger
- Copyright (C) 2010 Rudolf Marek r.marek@assembler.cz
- This program is free software; you can redistribute it and/or modify
- it under the terms of the GNU General Public License as published by
@@ -27,7 +28,77 @@ #include <errno.h> #include "flash.h"
-#ifdef __DARWIN__ +#ifdef __DJGPP__ [...] +#elif __DARWIN__
Can you change this to #elif defined(__DARWIN) to keep the logic as it was? I don't have a Mac OS X machine, so I can't test...
yes sorry fixed.
#include <DirectIO/darwinio.h>
#define MEM_DEV "DirectIO"
If you drop the board_enable.c hunk (and if my latest Makefile+README patch is OK for you), this is Acked-by: Carl-Daniel Hailfinger c-d.hailfinger.devel.2006@gmx.net
I can update the README once your pciutils patch has been merged.
I dont know who can commit this, I'm attaching the fixed patch.
Signed-off-by: Rudolf Marek r.marek@assembler.cz
Please check that in, so at least something is there ;) The libpci fixes etc can wait until Martin shows up.
Thanks, Rudolf
On 16.03.2010 23:40, Rudolf Marek wrote:
I added a short version of your description to the flashrom README.
Ok
I made some more changes to the Makefile. "make dos" is something I'd rather avoid because it hardcodes too many assumptions.
I would prefer to have there something like:
make djgpp-dos
because I will tend to forgot how to invoke the make from command line.
Index: physmap.c
--- physmap.c (revision 935) +++ physmap.c (working copy) @@ -4,6 +4,7 @@
- Copyright (C) 2009 Peter Stuge peter@stuge.se
- Copyright (C) 2009 coresystems GmbH
- Copyright (C) 2010 Carl-Daniel Hailfinger
- Copyright (C) 2010 Rudolf Marek r.marek@assembler.cz
- This program is free software; you can redistribute it and/or
modify
- it under the terms of the GNU General Public License as
published by
@@ -27,7 +28,77 @@ #include <errno.h> #include "flash.h"
-#ifdef __DARWIN__ +#ifdef __DJGPP__ [...] +#elif __DARWIN__
Can you change this to #elif defined(__DARWIN) to keep the logic as it was? I don't have a Mac OS X machine, so I can't test...
yes sorry fixed.
#include <DirectIO/darwinio.h>
#define MEM_DEV "DirectIO"
If you drop the board_enable.c hunk (and if my latest Makefile+README patch is OK for you), this is Acked-by: Carl-Daniel Hailfinger c-d.hailfinger.devel.2006@gmx.net
I can update the README once your pciutils patch has been merged.
I dont know who can commit this, I'm attaching the fixed patch.
Signed-off-by: Rudolf Marek r.marek@assembler.cz
Please check that in, so at least something is there ;) The libpci fixes etc can wait until Martin shows up.
Except for the Makefile (which needs more discussion) this is Acked-by: Carl-Daniel Hailfinger c-d.hailfinger.devel.2006@gmx.net
Thanks a lot for making this possible.
Regards, Carl-Daniel
On 17.03.2010 00:55, Carl-Daniel Hailfinger wrote:
On 16.03.2010 23:40, Rudolf Marek wrote:
I added a short version of your description to the flashrom README.
Ok
I made some more changes to the Makefile. "make dos" is something I'd rather avoid because it hardcodes too many assumptions.
I would prefer to have there something like:
make djgpp-dos
because I will tend to forgot how to invoke the make from command line.
Index: physmap.c
--- physmap.c (revision 935) +++ physmap.c (working copy) @@ -4,6 +4,7 @@
- Copyright (C) 2009 Peter Stuge peter@stuge.se
- Copyright (C) 2009 coresystems GmbH
- Copyright (C) 2010 Carl-Daniel Hailfinger
- Copyright (C) 2010 Rudolf Marek r.marek@assembler.cz
- This program is free software; you can redistribute it and/or
modify
- it under the terms of the GNU General Public License as
published by
@@ -27,7 +28,77 @@ #include <errno.h> #include "flash.h"
-#ifdef __DARWIN__ +#ifdef __DJGPP__ [...] +#elif __DARWIN__
Can you change this to #elif defined(__DARWIN) to keep the logic as it was? I don't have a Mac OS X machine, so I can't test...
yes sorry fixed.
#include <DirectIO/darwinio.h>
#define MEM_DEV "DirectIO"
If you drop the board_enable.c hunk (and if my latest Makefile+README patch is OK for you), this is Acked-by: Carl-Daniel Hailfinger c-d.hailfinger.devel.2006@gmx.net
I can update the README once your pciutils patch has been merged.
I dont know who can commit this, I'm attaching the fixed patch.
Signed-off-by: Rudolf Marek r.marek@assembler.cz
Please check that in, so at least something is there ;) The libpci fixes etc can wait until Martin shows up.
Except for the Makefile (which needs more discussion) this is Acked-by: Carl-Daniel Hailfinger c-d.hailfinger.devel.2006@gmx.net
Thanks a lot for making this possible.
And committed in r944.
I'll repost Makefile and README in a slightly updated version.
Regards, Carl-Daniel