HP Vectra VL420 SFF

[Mattias Mattsson]

Hi,

I'm trying to flash the latest BIOS-version on a HP Vectra VL420 SFF
but it fails with latest flashrom (r811).

This computer has a motherboard similiar to an ASUS P4B-MX, the HP
board is even labeled "P4B-MX" on the board itself. Some differences
though: The HP board uses an AMI BIOS while the ASUS seems to have an
Award BIOS.

Latest VL420-BIOS (version "JA.01.07US"):
ftp://ftp.hp.com/pub/softpaq/sp30001-30500/sp30221.exe

Info attached.

Thanks!
-mattias

[Michael Karcher]

Am Dienstag, den 22.12.2009, 21:05 +0100 schrieb Mattias Mattsson:
> I'm trying to flash the latest BIOS-version on a HP Vectra VL420 SFF
> but it fails with latest flashrom (r811).

This board needs a board enable. I don't like the PCI ID match, as HP
seems to use subsystem device IDs based on chip device IDs instead of
their products. The match is likely to broad.

Mattias, please try this patch, also, please send the output of
"dmidecode -s baseboard-manufacturer" and "dmidecode -s
baseboard-product-name" to this list, in case we are going to implement
DMI-based device detection. Your machine would be a candidate where it
seems useful.

Regards,
  Michael Karcher

Index: board_enable.c
===================================================================
--- board_enable.c	(Revision 803)
+++ board_enable.c	(Arbeitskopie)
@@ -830,7 +830,9 @@
 }

/**
- * Suited for ASUS P4B266: socket478 + intel 845D + ICH2.
+ * Suited for:
+ *  - Asus P4B266: socket478 + intel 845D + ICH2.
+ *  - (Asus?) P4B-MX (in HP Vectra VL420 SFF): socket 478 + 845D + ICH2
  */
 static int intel_ich_gpio22_raise(const char *name)
 {
@@ -1246,6 +1248,7 @@
    {0x1002, 0x4398, 0x1458, 0x5004,  0x1002, 0x4391, 0x1458, 0xb002, NULL,         NULL,          "GIGABYTE",    "GA-MA78GM-S2H",      it87xx_probe_spi_flash},
    {0x1002, 0x438d, 0x1458, 0x5001,  0x1002, 0x5956, 0x1002, 0x5956, NULL,         NULL,          "GIGABYTE",    "GA-MA790FX-DQ6",     it87xx_probe_spi_flash},
    {0x1166, 0x0223, 0x103c, 0x320d,  0x102b, 0x0522, 0x103c, 0x31fa, "hp",         "dl145_g3",    "HP",          "DL145 G3",           board_hp_dl145_g3_enable},
+	{0x8086, 0x1a30, 0x103c, 0x1a30,  0x8086, 0x2443, 0x103c, 0x2440, NULL,         NULL,          "HP",          "VL420 SFF",          intel_ich_gpio22_raise},
    {0x1166, 0x0205, 0x1014, 0x0347,  0x1002, 0x515E, 0x1014, 0x0325, NULL,         NULL,          "IBM",         "x3455",              board_ibm_x3455},
    {0x1039, 0x5513, 0x8086, 0xd61f,  0x1039, 0x6330, 0x8086, 0xd61f, NULL,         NULL,          "Intel",       "D201GLY",            wbsio_check_for_spi},
    {0x1022, 0x7468,      0,      0,       0,      0,      0,      0, "iwill",      "dk8_htx",     "IWILL",       "DK8-HTX",            w83627hf_gpio24_raise_2e},

[Mattias Mattsson]

On Wed, Dec 23, 2009 at 00:50, Michael Karcher
 wrote:
> Am Dienstag, den 22.12.2009, 21:05 +0100 schrieb Mattias Mattsson:
> > I'm trying to flash the latest BIOS-version on a HP Vectra VL420 SFF
> > but it fails with latest flashrom (r811).
>
> This board needs a board enable. I don't like the PCI ID match, as HP
> seems to use subsystem device IDs based on chip device IDs instead of
> their products. The match is likely to broad.
>
> Mattias, please try this patch, also, please send the output of
> "dmidecode -s baseboard-manufacturer" and "dmidecode -s
> baseboard-product-name" to this list, in case we are going to implement
> DMI-based device detection. Your machine would be a candidate where it
> seems useful.

The patch works. Many thanks!

Here's the requested output:

# dmidecode -s baseboard-manufacturer
Hewlett-Packard

# dmidecode -s baseboard-product-name
HP System Board

The above seems rather generic so here's also:

# dmidecode -s baseboard-version
A04

-mattias

[Michael Karcher]

Am Mittwoch, den 23.12.2009, 01:35 +0100 schrieb Mattias Mattsson:
> > Mattias, please try this patch, also, please send the output of
> > "dmidecode -s baseboard-manufacturer" and "dmidecode -s
> > baseboard-product-name" to this list, in case we are going to implement
> > DMI-based device detection. Your machine would be a candidate where it
> > seems useful.
>
> The patch works. Many thanks!

Thank you for your quick feedback.
> Here's the requested output:
>
> # dmidecode -s baseboard-manufacturer
> Hewlett-Packard
>
> # dmidecode -s baseboard-product-name
> HP System Board
>
> The above seems rather generic so here's also:
>
> # dmidecode -s baseboard-version
> A04

The version alone doesn't tell us much, and manufacturer HP can already
be determined by the PCI IDs. Can you check the full dmidecode output
for some string we can use to detect the VL420 SFF hardware? Maybe
system-product-name. If you are to attach a plain dmidecode output,
please remove all serial numbers/GUIDs.

Regards,
  Michael Karcher

[Mattias Mattsson]

> The version alone doesn't tell us much, and manufacturer HP can already
> be determined by the PCI IDs. Can you check the full dmidecode output
> for some string we can use to detect the VL420 SFF hardware? Maybe
> system-product-name. If you are to attach a plain dmidecode output,
> please remove all serial numbers/GUIDs.

The attached file contains full output from dmidecode.

Thanks again!
-mattias

[PATCH] Board enable for HP Vectra VL420SFF

[Michael Karcher]

This code has been tested by Mattias Mattsson on 23. December 2009,
but without the DMI match. Now that DMI support is in and working,
the board can be added to mainline flashrom.

Signed-off-by: Michael Karcher 
---
Hello Mattias, could you please try whether current svn flashrom (at least
r896) with this patch applied outputs
 Disabling write protection for HP VL420 SFF
on invocation with "flashrom -V"? If yes, please reply to this mail with a
mail containing the line "Acked-by: Mattias Mattsson ".
You don't have to try anything dangerous like erasing/writing the flash
chip, as only the board recognition has changed.

Regards,
  Michael Karcher

board_enable.c |    5 ++++-
 1 files changed, 4 insertions(+), 1 deletions(-)

diff --git a/board_enable.c b/board_enable.c
index 2de5fd7..2aa9c10 100644
--- a/board_enable.c
+++ b/board_enable.c
@@ -869,7 +869,9 @@ static int intel_ich_gpio21_raise(const char *name)
 }

/**
- * Suited for ASUS P4B266: socket478 + intel 845D + ICH2.
+ * Suited for:
+ *  - Asus P4B266: socket478 + intel 845D + ICH2.
+ *  - Asus P4B-MX variant in HP Vectra VL420 SFF: socket478 + 845D + ICH2
  */
 static int intel_ich_gpio22_raise(const char *name)
 {
@@ -1224,6 +1226,7 @@ struct board_pciid_enable board_pciid_enables[] = {
    {0x1002, 0x4398, 0x1458, 0x5004,  0x1002, 0x4391, 0x1458, 0xb002, NULL,          NULL,         NULL,          "GIGABYTE",    "GA-MA78GM-S2H",         0,   it87xx_probe_spi_flash},
    {0x1002, 0x438d, 0x1458, 0x5001,  0x1002, 0x5956, 0x1002, 0x5956, NULL,          NULL,         NULL,          "GIGABYTE",    "GA-MA790FX-DQ6",        0,   it87xx_probe_spi_flash},
    {0x1166, 0x0223, 0x103c, 0x320d,  0x102b, 0x0522, 0x103c, 0x31fa, NULL,          "hp",         "dl145_g3",    "HP",          "DL145 G3",              0,   board_hp_dl145_g3_enable},
+	{0x8086, 0x1a30, 0x103c, 0x1a30,  0x8086, 0x2443, 0x103c, 0x2440, "^VL420$",     NULL,         NULL,          "HP",          "VL420 SFF",             0,   intel_ich_gpio22_raise},
    {0x1166, 0x0205, 0x1014, 0x0347,  0x1002, 0x515E, 0x1014, 0x0325, NULL,          NULL,         NULL,          "IBM",         "x3455",                 0,   board_ibm_x3455},
    {0x1039, 0x5513, 0x8086, 0xd61f,  0x1039, 0x6330, 0x8086, 0xd61f, NULL,          NULL,         NULL,          "Intel",       "D201GLY",               0,   wbsio_check_for_spi},
    {0x1022, 0x7468,      0,      0,       0,      0,      0,      0, NULL,          "iwill",      "dk8_htx",     "IWILL",       "DK8-HTX",               0,   w83627hf_gpio24_raise_2e},