SeaBIOS February 2012

seabios@seabios.org

27 participants
67 discussions

Re: [SeaBIOS] [PATCH 1/3] vgahooks: Allow usage of VGA hooks in the Xen case
by Ian Campbell 02 Feb '12

02 Feb '12

On Thu, 2012-02-02 at 13:28 +0000, Julian Pidancet wrote: > On Wed, Feb 1, 2012 at 4:04 AM, Kevin O'Connor <kevin(a)koconnor.net> wrote: > > On Mon, Jan 30, 2012 at 01:55:54PM +0000, julian.pidancet(a)gmail.com wrote: > >> From: Julian Pidancet <julian.pidancet(a)citrix.com> > >> > >> This patch enables the VGAHOOKS config option if the BIOS is configured > >> for Xen. > >> In the Xen VGA passthrough case, the option ROM can be extracted from > >> the physical device by hvmloader and pre-deployed for seabios. This > >> patch also allows setting up VGA hooks in case the VGA option ROM is > >> already deployed by hvmloader. > > > > This seems a bit odd - if option roms are deployed, how can seabios > > reliably find the vga device? Wouldn't it be simpler to just use > > seabios in non-optionrom-deployed mode? (After all, wouldn't seabios > > still be able to pull the roms even if hvmloader alread did?) > > > > I think you're right, I just checked the sources of hvmloader in > xen-unstable, and it seems the option roms are only pre-deployed when > using rombios instead of seabios. That's correct. When Xen uses SeaBIOS we let it take care of the option ROMs. Ian. -- Ian Campbell "All Bibles are man-made." -- Thomas Edison

1 0

Re: [SeaBIOS] [PATCH 2/3] vgahooks: Register 155f vga hooks for Xen Intel passthrough devices
by Kevin O'Connor 02 Feb '12

02 Feb '12

On Mon, Jan 30, 2012 at 01:55:55PM +0000, julian.pidancet(a)gmail.com wrote: > From: Julian Pidancet <julian.pidancet(a)citrix.com> > > Some Intel VGA option ROMs require some 155f hooks implemented in the > BIOS. > > Signed-off-by: Julian Pidancet <julian.pidancet(a)citrix.com> > --- > src/vgahooks.c | 20 +++++++++++++++++++- > 1 files changed, 19 insertions(+), 1 deletions(-) > > diff --git a/src/vgahooks.c b/src/vgahooks.c > index a8f667c..ca12e31 100644 > --- a/src/vgahooks.c > +++ b/src/vgahooks.c > @@ -11,6 +11,7 @@ > #include "pci_ids.h" // PCI_VENDOR_ID_VIA > #include "util.h" // handle_155f > #include "config.h" // CONFIG_* > +#include "xen.h" // usingXen > > #define VH_VIA 1 > #define VH_INTEL 2 > @@ -228,6 +229,15 @@ getac_setup(struct pci_device *pci) > { > } > > +static void > +xen_vgapt_setup(struct pci_device *pci) > +{ > + if (pci->vendor == 0x8086 && pci->class == 0x300) { > + VGAHookHandlerType = VH_INTEL; > + IntelDisplayType = BOOT_DISPLAY_DEFAULT; > + IntelDisplayId = 3; > + } > +} > > /**************************************************************** > * Entry and setup > @@ -254,7 +264,15 @@ handle_155f(struct bregs *regs) > void > vgahook_setup(struct pci_device *pci) > { > - if (!CONFIG_VGAHOOKS || !CBvendor || !CBpart) > + if (!CONFIG_VGAHOOKS) > + return; > + > + if (usingXen()) { > + xen_vgapt_setup(pci); > + return; > + } > + > + if (!CBvendor || !CBpart) > return; Is Xen passthrough special, or will the same thing occur on KVM passthrough? If the 155f hooks are needed on Intel VGA cards, I wouldn't tie it to Xen - we can just turn it on by default and enable it if the VGA card is recognized as Intel based. -Kevin

2 1

Re: [SeaBIOS] [PATCH 1/3] vgahooks: Allow usage of VGA hooks in the Xen case
by Kevin O'Connor 02 Feb '12

02 Feb '12

On Mon, Jan 30, 2012 at 01:55:54PM +0000, julian.pidancet(a)gmail.com wrote: > From: Julian Pidancet <julian.pidancet(a)citrix.com> > > This patch enables the VGAHOOKS config option if the BIOS is configured > for Xen. > In the Xen VGA passthrough case, the option ROM can be extracted from > the physical device by hvmloader and pre-deployed for seabios. This > patch also allows setting up VGA hooks in case the VGA option ROM is > already deployed by hvmloader. This seems a bit odd - if option roms are deployed, how can seabios reliably find the vga device? Wouldn't it be simpler to just use seabios in non-optionrom-deployed mode? (After all, wouldn't seabios still be able to pull the roms even if hvmloader alread did?) -Kevin

2 1

[PATCH 0/6] More vgabios refactoring
by Kevin O'Connor 01 Feb '12

01 Feb '12

Some additional changes. Things should be about to wind down on the vgabios front. Cirrus should be about done. For bochsvga, only vbe functions 04/08 remain to be ported. (Per Gerd's feedback, I don't think it's worth porting the 32bit vbe interface.) -Kevin Kevin O'Connor (6): vgabios: Compare PCI ids against pci rom struct instead of config settings. vgabios: Fix linelength calculations in bochsvga and vbe. vgabios: Rework bochsvga mode switching. vgabios: Simplify cirrus find mode code. vgabios: Minor - organize cirrus code so like functions are near each other. vgabios: Move bocshvga mode checking from runtime to init. vgasrc/bochsvga.c | 206 +++++++++++++++++------------------------- vgasrc/bochsvga.h | 4 +- vgasrc/clext.c | 255 +++++++++++++++++++++------------------------------- vgasrc/clext.h | 4 +- vgasrc/vbe.c | 13 +-- vgasrc/vgabios.c | 6 +- 6 files changed, 199 insertions(+), 289 deletions(-) -- 1.7.6.4

1 6

Re: [SeaBIOS] [PATCH v2] Introduce config option to select which IO port to send debug output on
by Kevin O'Connor 01 Feb '12

01 Feb '12

On Wed, Feb 01, 2012 at 04:03:24PM +0000, julian.pidancet(a)gmail.com wrote: > From: Julian Pidancet <julian.pidancet(a)citrix.com> > > Introduce CONFIG_DEBUG_IO_PORT config option which allow user > to choose on which port to send debug on at configure time. > > Bochs users are likely to use 0x402 (default) for debugging purposes, > whereas Xen uses 0xe9, as IO port address. Thanks - I've added it to my queue. -Kevin

1 0

[PATCH v2] Introduce config option to select which IO port to send debug output on
by julian.pidancet＠gmail.com 01 Feb '12

01 Feb '12

From: Julian Pidancet <julian.pidancet(a)citrix.com> Introduce CONFIG_DEBUG_IO_PORT config option which allow user to choose on which port to send debug on at configure time. Bochs users are likely to use 0x402 (default) for debugging purposes, whereas Xen uses 0xe9, as IO port address. v2: Typo. Signed-off-by: Julian Pidancet <julian.pidancet(a)citrix.com> --- src/Kconfig | 18 ++++++++++++++++++ src/output.c | 4 ++-- 2 files changed, 20 insertions(+), 2 deletions(-) diff --git a/src/Kconfig b/src/Kconfig index 9ea48bb..f57899b 100644 --- a/src/Kconfig +++ b/src/Kconfig @@ -348,4 +348,22 @@ menu "Debugging" default 0x3f8 help Base port for serial - generally 0x3f8, 0x2f8, 0x3e8, or 0x2e8. + + config DEBUG_IO + depends on !COREBOOT && DEBUG_LEVEL != 0 + bool "Special IO port debugging" + default y + help + Some emulators or hypervisors provide with a way to output debug + information by outputing strings in a special port present in the + IO space. + + config DEBUG_IO_PORT + depends on DEBUG_IO + hex "Debug IO port address" + default 0x0402 + help + Bochs uses the 0x0402 address by default, whereas Xen + makes the 0xe9 IO address available for guests use. + endmenu diff --git a/src/output.c b/src/output.c index 462ffb7..bdde7cc 100644 --- a/src/output.c +++ b/src/output.c @@ -75,9 +75,9 @@ putc_debug(struct putcinfo *action, char c) { if (! CONFIG_DEBUG_LEVEL) return; - if (! CONFIG_COREBOOT) + if (CONFIG_DEBUG_IO) // Send character to debug port. - outb(c, PORT_BIOS_DEBUG); + outb(c, CONFIG_DEBUG_IO_PORT); if (c == '\n') debug_serial('\r'); debug_serial(c); -- Julian Pidancet

1 0

Re: [SeaBIOS] [PATCH] Introduce config option to select which IO port to send debug output on
by Kevin O'Connor 01 Feb '12

01 Feb '12

On Mon, Jan 30, 2012 at 05:23:19PM +0000, julian.pidancet(a)gmail.com wrote: > --- a/src/output.c > +++ b/src/output.c > @@ -75,9 +75,9 @@ putc_debug(struct putcinfo *action, char c) > { > if (! CONFIG_DEBUG_LEVEL) > return; > - if (! CONFIG_COREBOOT) > + if (! CONFIG_DEBUG_IO) Inverted test? -Kevin > // Send character to debug port. > - outb(c, PORT_BIOS_DEBUG); > + outb(c, CONFIG_DEBUG_IO_PORT); > if (c == '\n') > debug_serial('\r'); > debug_serial(c);

2 1

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

SeaBIOS February 2012