On 01/23/13 00:03, David Woodhouse wrote:
> Without CSM support, I can reproduce the problem with the 'Press any key
> to boot from CD or DVD...' message not being displayed, but *only* if I
> start qemu with '-vga vmware'. With std, qxl or cirrus options, OVMF now
> works fine. I think it has nati
On 01/22/13 14:46, David Woodhouse wrote:
> On Tue, 2013-01-22 at 09:51 +0100, Gerd Hoffmann wrote:
>>
>>> And we only invoke the iPXE ROM on the network card *after* the user has
>>> seen the boot menu (without that option), so the user never gets to
>>> choose that.
>>
>> You can build ipxe efi r
On Tue, 2013-01-22 at 17:29 +0100, Laszlo Ersek wrote:
>
> Does this mean the VGA BIOS / int 0x10 are available even to a UEFI OS
> being booted directly from a UEFI boot option?
>
> If so, perhaps Jeff might want to retry installing windows 2008 R2sp1 on
> OVMF (= a build that includes the SeaBI
On Tue, 2013-01-22 at 09:51 -0800, Andrew Fish wrote:
>
> So the history here is:
...
> So some of the Intel Framework stuff did not make it into the PI
> specification. Most notably all the CSM stuff. So that is why the edk2
> has the IntelFramworkPkg.
>
> So in general when I say EFI I mean t
On Jan 22, 2013, at 10:51 AM, David Woodhouse wrote:
> On Tue, 2013-01-22 at 09:51 -0800, Andrew Fish wrote:
>> Basically anything in generic EFI code that assumes a CSM (legacy
>> BIOS) exists is a bug.
>
> Yes. But checking for the existence of it and using it *if* it's
> available would not
On Tue, 2013-01-22 at 09:51 -0800, Andrew Fish wrote:
> Basically anything in generic EFI code that assumes a CSM (legacy
> BIOS) exists is a bug.
Yes. But checking for the existence of it and using it *if* it's
available would not be a bug.
> Thus you would want to try an PXE boot EFI and if th
On Jan 22, 2013, at 5:46 AM, David Woodhouse wrote:
> Thinking about netboot... sometimes I might be booting a legacy OS from
> the network, and sometimes I might be booting an EFI-aware OS. Do I have
> to use the *right* version of the PXE ROM, according to which OS it's
> going to find on the
On 01/22/13 18:10, Andrew Fish wrote:
>> Does this mean the VGA BIOS / int 0x10 are available even to a UEFI OS
>> being booted directly from a UEFI boot option?
>>
>
> NO!
I fully agree that we should make no such promise (I vaguely recall one
spec or another explicitly warning about this, ie.
On Tue, 2013-01-22 at 09:10 -0800, Andrew Fish wrote:
> > Does this mean the VGA BIOS / int 0x10 are available even to a UEFI OS
> > being booted directly from a UEFI boot option?
> >
>
> NO!
I think you're answering a different question to the one that was asked.
We all know that a sane operat
On Tue, 2013-01-22 at 17:29 +0100, Laszlo Ersek wrote:
> Does this mean the VGA BIOS / int 0x10 are available even to a UEFI OS
> being booted directly from a UEFI boot option?
I believe so, yes. OVMF even has a BiosVideo driver which will use the
INT 10h services for display.
> If so, perhaps Je
On Jan 22, 2013, at 8:29 AM, Laszlo Ersek wrote:
> On 01/22/13 14:46, David Woodhouse wrote:
>
>> I have the CSM working reasonably well now for basic functionality. IDE
>> disks and CDROMs show up in the EFI boot menu for booting in legacy
>> mode, and if you choose them then SeaBIOS will boot
On 01/22/13 14:46, David Woodhouse wrote:
> I have the CSM working reasonably well now for basic functionality. IDE
> disks and CDROMs show up in the EFI boot menu for booting in legacy
> mode, and if you choose them then SeaBIOS will boot what it's asked to
> boot. Windows, Linux and Solaris 11
On Tue, 2013-01-22 at 09:51 +0100, Gerd Hoffmann wrote:
>
> > And we only invoke the iPXE ROM on the network card *after* the user has
> > seen the boot menu (without that option), so the user never gets to
> > choose that.
>
> You can build ipxe efi roms and use those, then netboot will show up
Hi,
> And we only invoke the iPXE ROM on the network card *after* the user has
> seen the boot menu (without that option), so the user never gets to
> choose that.
You can build ipxe efi roms and use those, then netboot will show up in
the efi boot menu (although I didn't got that actually work
On Jan 21, 2013, at 4:43 PM, David Woodhouse wrote:
> On Tue, 2013-01-22 at 00:29 +, Li, Elvin wrote:
>> You can make a try to implement such CSM16 to return back to EFI,
>> GenericLegacyBoot()
>> thus returns EFI_DEVICE_UNSUPPORTED, but it is BDS which initially
>> calls "LegacyBios->Legacy
On Tue, 2013-01-22 at 00:29 +, Li, Elvin wrote:
> You can make a try to implement such CSM16 to return back to EFI,
> GenericLegacyBoot()
> thus returns EFI_DEVICE_UNSUPPORTED, but it is BDS which initially
> calls "LegacyBios->LegacyBoot ()" to launch legacy boot, you can add
> error handling
I drivers and go to next boot
options.
Thanks
Elvin
-Original Message-
From: David Woodhouse [mailto:dw...@infradead.org]
Sent: Monday, January 21, 2013 12:00 PM
To: Andrew Fish
Cc: edk2-devel@lists.sourceforge.net
Subject: Re: [edk2] CSM and keyboard handling
On Sun, 2013-01-20 a
On Sun, 2013-01-20 at 19:54 -0800, Andrew Fish wrote:
> The basic problem is as soon as the CSM calls something outside the
> CSM code, then you don't know what kind of memory has been overwritten
> or hardware written to. So the design was always it is not safe to
> come back, as there is not a cl
On Jan 20, 2013, at 5:53 PM, "Li, Elvin" wrote:
> If it matters, bear in mind that most of my CSM calls for now are just
> dummies that do almost nothing except return success. The only thing my
> CSM really does is populate the interrupt vectors (INT 15h is needed
> even to *return* from the Leg
On Mon, 2013-01-21 at 01:53 +, Li, Elvin wrote:
> [Li, Elvin] It is unlucky that CSM32 does not support returning back
> from legacy boot at all. If legacy boot is failed, a full function of
> CSM16 normally keep in legacy environment and INT19 handler indicates
> “No bootable device -- insert
David:
Some comments are added below. Search [Li, Elvin].
Thanks
Elvin
-Original Message-
From: David Woodhouse [mailto:dw...@infradead.org]
Sent: Friday, January 18, 2013 8:44 PM
To: Li, Elvin
Cc: edk2-devel@lists.sourceforge.net
Subject: Re: [edk2] CSM and keyboard handling
On Fri, 2013-01-18 at 02:59 +, Li, Elvin wrote:
> Hi David:
> CSM framework has mentioned the side effect that "all EFI
> drivers are disconnected and must be reconnected for proper EFI
> functioning". So if you call the legacy bios protocol interface
> LegacyBiosShadowAllLegacyOproms w
[edk2] CSM and keyboard handling
In PciShadowRoms() in LegacyBiosDxe/LegacyPci.c we disconnect the EFI keyboard
driver before attempting to invoke any option roms.
After that point, what is supposed to provide keyboard service? There's a
BiosKeyboard driver but that isn't enabled and
In PciShadowRoms() in LegacyBiosDxe/LegacyPci.c we disconnect the EFI
keyboard driver before attempting to invoke any option roms.
After that point, what is supposed to provide keyboard service? There's
a BiosKeyboard driver but that isn't enabled and built in OVMF, even if
CSM_ENABLE is defined.
24 matches
Mail list logo
