Re: [coreboot] Fwd: [PATCH v3 1/1] libpayload: add support for finding and parsing Coreboot BIOS tables
On Sun, 21 Aug 2011 17:01:28 -0700 Philip Prindeville philipp@redfish-solutions.com wrote:
Attached.
oookay. I'll await a formal patch with all the cc's, etc.
As for the "upstream" version: if/when this patch is merged, it *is* the upstream version. The coreboot tree will need to drop the old version and migrate to the new.
Please be sure to carefully describe the background to the code within the final changelog. Lots of people won't even know what coreboot *is*. Spell this out completely and be sure to tell the kernel developers about the value of this patch to our users.
On Mon, 22 Aug 2011 15:01:51 -0700 Andrew Morton akpm@linux-foundation.org wrote:
On Sun, 21 Aug 2011 17:01:28 -0700 Philip Prindeville philipp@redfish-solutions.com wrote:
Attached.
oookay. I'll await a formal patch with all the cc's, etc.
Also, if you email the patch inline (rather than as an attachment), then we can include comments inline as well.
As for the "upstream" version: if/when this patch is merged, it *is* the upstream version. The coreboot tree will need to drop the old version and migrate to the new.
IOW, make sure the coreboot folks are okay with what's being submitted as well!
Please be sure to carefully describe the background to the code within the final changelog. Lots of people won't even know what coreboot *is*. Spell this out completely and be sure to tell the kernel developers about the value of this patch to our users.
I'd second this. I'm minimally familiar with coreboot, but I have no idea what the purpose behind coreboot BIOS tables are.
On 8/22/11 3:01 PM, Andrew Morton wrote:
On Sun, 21 Aug 2011 17:01:28 -0700 Philip Prindeville philipp@redfish-solutions.com wrote:
Attached.
oookay. I'll await a formal patch with all the cc's, etc.
As for the "upstream" version: if/when this patch is merged, it *is* the upstream version. The coreboot tree will need to drop the old version and migrate to the new.
So that we're on the same page, I took a different version of the code that is normally linked into standalone ELF images that run directly from the bootloader, much as (say) memtest86+ does.
I took that standalone library and ported it to run inside the linux kernel, adding printk()'s, phys_to_virt(), and ioremap() calls to make it work.
What I meant was I was hoping that some of the other changes made can be ported back into the arch/x86 version (like the using void * pointers instead of unsigned char * and then casting to struct pointers).
Please be sure to carefully describe the background to the code within the final changelog. Lots of people won't even know what coreboot *is*. Spell this out completely and be sure to tell the kernel developers about the value of this patch to our users.
Good point.
Thanks.
On 8/22/11 4:00 PM, Andres Salomon wrote:
On Mon, 22 Aug 2011 15:01:51 -0700 Andrew Morton akpm@linux-foundation.org wrote:
On Sun, 21 Aug 2011 17:01:28 -0700 Philip Prindeville philipp@redfish-solutions.com wrote:
Attached.
oookay. I'll await a formal patch with all the cc's, etc.
Also, if you email the patch inline (rather than as an attachment), then we can include comments inline as well.
I wish I could, but alas I use Thunderbird. Some misguided hacker rewrote the HTML tree-parser and now it generates broken text/plain as a consequence.
Sending attachments is the only way to stop it from mangling.
Yes, I know... I could run another mail agent, but I'm also a contributor to Thunderbird (though in a different area than the HTML parsing) so I try to eat my own dogfood when I can.
As for the "upstream" version: if/when this patch is merged, it *is* the upstream version. The coreboot tree will need to drop the old version and migrate to the new.
IOW, make sure the coreboot folks are okay with what's being submitted as well!
Been trying to but the mailing list and IRC channel have been challenging to evoke any sort of response from.
Please be sure to carefully describe the background to the code within the final changelog. Lots of people won't even know what coreboot *is*. Spell this out completely and be sure to tell the kernel developers about the value of this patch to our users.
I'd second this. I'm minimally familiar with coreboot, but I have no idea what the purpose behind coreboot BIOS tables are.
It allows me to identify the manufacturer and board model by querying BIOS for that.
Necessary for platform drivers that don't have another way of detecting the hardware they are running on.
Am 23.08.2011 07:53 schrieb Philip Prindeville:
On 8/22/11 4:00 PM, Andres Salomon wrote:
Also, if you email the patch inline (rather than as an attachment), then we can include comments inline as well.
I wish I could, but alas I use Thunderbird. Some misguided hacker rewrote the HTML tree-parser and now it generates broken text/plain as a consequence.
Sending attachments is the only way to stop it from mangling.
Yes, I know... I could run another mail agent, but I'm also a contributor to Thunderbird (though in a different area than the HTML parsing) so I try to eat my own dogfood when I can.
The following trick works for me in newer Seamonkey which has the same bug: mailnews.send_plaintext_flowed=false
Regards, Carl-Daniel
-
Andres Salomon -
Andrew Morton -
Carl-Daniel Hailfinger -
Philip Prindeville