Kevin O'Connor [Mon, 21 Jan 2013 16:38:49 +0000 (11:38 -0500)]
POST: Reorganize post entry and "preinit" functions.
Unlocking ram in handle_post() is tricky and only needed under qemu.
Separate out that logic from the coreboot/xen paths by invoking
handle_elf_post separately. This simplifies both the qemu and
non-qemu code paths.
Also, organize all the "pre-init" functions into one section of the
file.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Kevin O'Connor [Wed, 6 Feb 2013 03:36:13 +0000 (22:36 -0500)]
Ensure exported symbols are visible in the final link
I'm about to introduce some post-processing in checkrom.py which will want
access to public symbols. So let's make sure they're defined in the final
link even if they're *not* cross-referenced from a different code section.
build: Define 16bit exported symbols in final rom.o object.
Based on patch by: David Woodhouse <David.Woodhouse@intel.com> Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
David Woodhouse [Thu, 17 Jan 2013 21:23:46 +0000 (21:23 +0000)]
Complete and checksum EFI_COMPATIBILITY16_TABLE at build time
GCC, for reasons unknown, will refuse to compile code such as
extern int bar;
uint16_t foo = &bar;
The assembler would happily emit a R_386_16 relocation for this if asked
nicely, and all would be well. But instead, GCC complains about the
initialiser not being constant.
So we tend to fill in 16-bit offsets at run-time, which is only
moderately inefficient. But for the CSM table used by EFI, it doesn't
work. We need the offset to be present in the *image*, before a line of
our own code has been run. Likewise the checksum.
This special-cases the table and entry point in checkrom.py rather than
attempting to do something generic. I did have a functional generic
implementation which could be invoked from csm.c along the lines of
CHECKSUM(csm_compat_table, TableChecksum)
... and which would emit the required offsetof(typeof(\1), \2) into
a special data section which was elided from the final build but parsed
by 'objdump -s' and the location of the table to be checksummed was
inferred from the name of the *variable* that got put into that special
section... seriously, it's better just to special-case it.
It was baroque enough just for the checksums, and filling in the entry
point which required access to *two* symbols was probably going to involve
emitting a *string* into that special build-data section. It had to die.
So yes, we have hard-coded symbol names, and even magic numbers in the
python script for table offsets etc., but that's because this is an ABI.
It doesn't change, and neither do any of the other tables that we might
now consider filling at build time just to avoid having to spend time
on doing so at runtime.
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
David Woodhouse [Sat, 26 Jan 2013 01:33:58 +0000 (19:33 -0600)]
Export copy_smbios() from biostables.c
Yes, copy_table() would invoke it, but CSM will *only* use it for SMBIOS
tables and however much we trick the compiler by gratuitously checking
the table signature right before calling copy_table(), it still doesn't
seem to notice that fact. And emits code for all the other three cases
we don't care about.
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
David Woodhouse [Tue, 5 Feb 2013 16:01:31 +0000 (16:01 +0000)]
Clean up #if in pirtable.c. CONFIG_PIRTABLE can't be set if CONFIG_COREBOOT is
CONFIG_PIRTABLE, along with various other BIOS tables, is only selectable
if CONFIG_QEMU is selected. QEMU/COREBOOT/CSM is a choice, so you can't
enable CONFIG_PIRTABLE if you've selected COREBOOT or CSM. So kill the
excessive #if check for COREBOOT.
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Kevin O'Connor [Mon, 21 Jan 2013 06:53:31 +0000 (01:53 -0500)]
POST: Reorganize post init functions for better grouping and reusability.
Place the "interface initialization" functions together, "platform
hardware initialization" functions together, and "prepare to boot"
functions together. This may also be useful for using SeaBIOS as a
CSM.
This slightly changes the order of some function invocations, but
should otherwise not change code behavior at all.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Kevin O'Connor [Mon, 21 Jan 2013 06:14:12 +0000 (01:14 -0500)]
Normalize POST initialization function name suffixes.
The POST phase has to invoke many initialization functions, and these
functions can have complex inter-dependencies. Try to categorize the
functions into 4 classes:
preinit - functions called very early in POST where function ordering
is very important and the code has limited access to other
interfaces.
init - functions that initialize internal interfaces and standard
external interfaces. This code is generally not dependent on
particular hardware and typically does not communicate directly
with any hardware devices.
setup - functions which access hardware or are dependent on particular
hardware or platform devices.
prepboot - functions that finalize internal interfaces and that
prepare for the boot phase.
This patch attempts to normalize the suffixes - functions that used
_init(), _setup(), _finalize(), or similar that did not follow the
above pattern were renamed. Other than function name changes, there
should be no code impact to this patch.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Kevin O'Connor [Sun, 20 Jan 2013 15:43:54 +0000 (10:43 -0500)]
pmm: Use 'struct segoff_s' in pmm header.
Minor cleanup - use 'struct segoff_s' instead of a manual
segment/offset pair.
Also, while at it, define the header signature at compile time instead
of at run-time. (Defining at run-time is a hold over from when
variable changes were persistent across soft-reboots.)
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Paolo Bonzini [Thu, 10 Jan 2013 12:41:36 +0000 (13:41 +0100)]
vgabios: implement AX=1120H..1124H functions
These function only have to set INT 1Fh and INT 43h, and set
the BDA height + number of rows.
I could not find out whether AX=1120h should also set the character
height to 8. I think not, because INT 43h might still point to
14- or 16-pixel high characters and in this case INT 1Fh will not
be used at all. Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Kevin O'Connor [Sat, 12 Jan 2013 21:36:11 +0000 (16:36 -0500)]
boot: Support "halt" in the boot order to prevent default boot attempts.
Current seabios will attempt to boot from both configured devices and
a list of default devices. Sometimes it is desirable to only attempt
to boot from the configured devices and not from any of the default
devices. This patch adds support for a "HALT" item in the "bootorder"
file - its presence will prevent SeaBIOS from attempting any default
boot actions.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net> Signed-off-by: Amos Kong <akong@redhat.com>
Kevin O'Connor [Sat, 22 Dec 2012 03:52:50 +0000 (22:52 -0500)]
Minor: Separate UUID display from F12 boot prompt.
Place the extra newline as part of the F12 prompt instead of the
SeaBIOS version banner so that a UUID report (when present) is still
visually separated from the boot menu prompt.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Laszlo Ersek [Tue, 18 Dec 2012 04:11:39 +0000 (05:11 +0100)]
display_uuid(): fix incomplete check after the loop
In the v2->v3 change of what would become commit 37676f83
<http://www.seabios.org/pipermail/seabios/2012-December/005166.html>, the
defense against an initial "addr > end" condition ("wraparound") was
erroneously loosened.
Paolo Bonzini [Thu, 13 Dec 2012 07:42:02 +0000 (08:42 +0100)]
acpi: reintroduce LNKS
Commit 4540409 (acpi: LNKS is not needed, 2012-08-07) removed LNKS because
it basically worked by chance: _CRS returns something else then one of the
possible resources from _PRS, _DIS would not really disable the interrupt,
and there was no _SRS method. It just happened to work because all OSes
have some kind of special-casing for SCI.
Unfortunately, the code after the patch is also against the spec, and it
breaks FreeBSD because it treats IRQ 9 polarity as active low without
the Interrupt() entry. Actually, numeric _PRT entries are handled the
same in Linux and FreeBSD (as active-low). However, under Linux it just
happens to trigger another special casing of SCI which sets SCI up from
its override entry in the MADT, ignoring the DSDT completely.
This patch adds back the LNKS, but without using the PIIX register for
LNKA in its methods.
Tested-by: Luigi Rizzo <rizzo@iet.unipi.it> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Kevin O'Connor [Wed, 12 Dec 2012 03:40:16 +0000 (22:40 -0500)]
Add compile checking phase to build.
The build textually includes all the code together in one unit for the
compiler in order to use gcc's -fwhole-program option. Textually
including all the code can mask subtle code errors (eg, forgetting to
include the correct headers) and can lead to confusing error reports
from the compiler. So, compile each file individually in addition to
the normal textually inclusive build. This improves the error
checking while still obtaining the benefits of -fwhole-program.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Kevin O'Connor [Wed, 12 Dec 2012 02:59:45 +0000 (21:59 -0500)]
Don't use gcc's -combine option.
The combine option has been deprecated from gcc for the last few
revisions. The option doesn't add any real benefits, so there's no
real point in checking for it on the few systems that will still have
it.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Kevin O'Connor [Sun, 2 Dec 2012 07:33:33 +0000 (02:33 -0500)]
acpi: PCI routing DSDT simplifications.
Use define_link() macro in PIIX4 hardware definitions. Now that all
remaining irq links are the same it's possible to just use a macro to
define the link.
Port several size optimizations to the q35 irq routing definitions:
define the PRQx fields in the _SB scope to reduce size of register
references, define CRS and STA methods in _SB scope to reduce code
duplication, simplify acpi code.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Kevin O'Connor [Sun, 2 Dec 2012 06:15:03 +0000 (01:15 -0500)]
ACPI: Remove unnecessary Return(One) calls from q35 DSDT.
Recent iasl compilers warn about spurious return calls. Remove extra
return calls from the GPE event methods. This is the q35 dsdt
equivalent of patch 475ce3f5f6a986877d637ac39f6ee23484ecc992.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Kevin O'Connor [Sun, 2 Dec 2012 06:09:17 +0000 (01:09 -0500)]
acpi: Reindent DSDT code.
This patch is purely cosmetic - no code changes should be observed.
Use a consistent indentation style with the ACPI DSDT code:
1 - Use spaces (no tabs) and 4 space indentation
2 - Place opening braces on same line as statement declaration
3 - Don't put a space between statement and opening parenthesis,
except for control statements (If, Else, While, Return) where a
space is always present
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Isaku Yamahata [Wed, 28 Nov 2012 09:17:36 +0000 (10:17 +0100)]
seabios: q35: add dsdt
add dsdt for q35 chipset of qemu.
[jbaron: remove suspd bits since they are now auto-generated, move pci window to 0xb0000000,
add framework for auto generated pci windows]
[kraxel: remove pci express root ports + pci bridge]
Cc: Matthew Garrett <mjg59@srcf.ucam.org> Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp> Signed-off-by: Jason Baron <jbaron@redhat.com>
Jason Baron [Wed, 28 Nov 2012 09:17:30 +0000 (10:17 +0100)]
seabios: make mttr UC area setup dynamic
Set up the UC area of mtrr dynamically based on mtrr_base. This allows
the bios to work for other chipsets that might want to set the mtrr.
Since BUILD_MAX_HIGHMEM is no longer used we can remove the config parameter.
This change reverses the order of pci_setup() and smm_init() with
mtrr_setup().
Kevin O'Connor [Mon, 26 Nov 2012 16:18:11 +0000 (11:18 -0500)]
floppy: Allow floppy code to be used with coreboot.
Allow coreboot users to enable the floppy support (via CBFS files
etc/floppy0 and etc/floppy1) on real hardware. It is unknown if the
Bochs derived floppy controller support will work on real hardware,
but this option will make it easier for those wishing to test.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Kevin O'Connor [Mon, 26 Nov 2012 16:05:32 +0000 (11:05 -0500)]
pciinit: Align start of PCI memory on i440 chipset.
Align the start of PCI memory space to simplify mtrr ranges. This is
configured per chipset so future chipsets can configure different
start addresses.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Kevin O'Connor [Wed, 21 Nov 2012 01:29:26 +0000 (20:29 -0500)]
Collapse PCI device tables in pci_bios_init_devices().
Use just one "struct pci_device_id" device table list in the
pci_bios_init_devices initialization code instead of three separate
tables. There isn't any compelling reason to use multiple tables.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Hannes Reinecke [Tue, 13 Nov 2012 14:03:31 +0000 (15:03 +0100)]
megasas: Add boot support for LSI MegaRAID SAS
This patch adds boot support for LSI MegaRAID SAS controllers.
Signed-off-by: Hannes Reinecke <hare@suse.de> Cc: Kevin O'Connor <kevin@koconnor.net> Cc: Gerd Hofmann <kraxel@redhat.com> Cc: Alex Graf <agraf@suse.de> Cc: Paolo Bonzini <pbonzini@redhat.com>
Updated _CRS method for HPET, bringing it in line with the way it is
presented on recent hardware (e.g. Dell Latitude D630, MacPro5,1, etc);
Allows it to be detected and utilized from Mac OS X; Also tested OK on
Linux (F16 64-bit install DVD) and Windows (Win7 64-bit install DVD).
Bruce Rogers [Thu, 25 Oct 2012 19:48:16 +0000 (13:48 -0600)]
Provide correct dpte pointer to indicate absence of dpte
The Pheonix EDD v3.0 spec says that if no dpte is present,
the dpte pointer should contain ffffh:ffffh, not the 0h:0h
which is currently used. This bit me when using NetWare as
a KVM guest.
geodevga: Add configuration option for output modes
Geode GX2 and LX are supporting Flat Panels. Add new
configuration options to support different output modes.
You can choose between:
* CRT
* Flat Panel
* CRT and Flat Panel
Signed-off-by: Christian Gmeiner <christian.gmeiner@gmail.com>
The Flat Panel Display Controller belongs to the Video
Processor unit but its register are starting at offset
0x400. Provide functions to work with fp register.
Signed-off-by: Christian Gmeiner <christian.gmeiner@gmail.com>
This patch makes seabios use the acpi pmtimer instead of tsc for
timekeeping. The pmtimer has a fixed frequency and doesn't need
calibration, thus it doesn't suffer from calibration errors due to a
loaded host machine.
As all supported graphic adapters are supporting VGA we should
list them in int10 (AX = 4F00h). Also clext.c and bochsvga.c
are haveing a function named like is_XXX_mode() which checks
if its a default VGA mode or a VESA mode. If its a normal VGA
mode stdvga_set_mode() gets called.
Signed-off-by: Christian Gmeiner <christian.gmeiner@gmail.com>
Kevin O'Connor [Tue, 4 Sep 2012 17:16:36 +0000 (13:16 -0400)]
vgabios: Check for relocatable windows before advertising support for it.
Use the VBE_win_granularity field to determine if the memory windows
are relocatable. (Bochs and Cirrus vga support relocatable windows
while the standard vga controller does not.)
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Kevin O'Connor [Mon, 3 Sep 2012 17:54:28 +0000 (13:54 -0400)]
vgabios: Ignore mode flags in vbe get_mode_info call.
Ignore any mode flags a caller may have set on the vbe get_mode_info
call. The spec doesn't require ignoring of flags, but it appears at
least some real-world vgabios vbe implementations do this.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Paolo Bonzini [Tue, 7 Aug 2012 13:01:29 +0000 (15:01 +0200)]
acpi: LNKS is not needed
LNKS is a bit strange in that it reuses the same PIIX register as LNKA,
but has a different interrupt. This means that the _CRS it returns will
not be one of the possible resources from _PRS. This shows up in the
Linux boot logs as
ACPI: PCI Interrupt Link [LNKS] (IRQs 9) *0
Instead of that, we can simply use a hardwired interrupt index.
Paolo Bonzini [Thu, 2 Aug 2012 13:07:26 +0000 (15:07 +0200)]
acpi: build PCI hotplug devices from a single template
More than 1kb of data is taken by the 32 copies of the PCI hotplug SSDT
methods. We can build them from a single template like we do for CPUs
(wrapped in a Scope(\_SB.PCI0) block).
Three items differ for each slot: the device name, bits 16-23 of _ADR,
the _SUN value. On top of this we have to rename the eject method for
non-removable slots, like we already do in build_pcihp.
There is a small change in the ASL: instead of including the number of
the slot in the implementation of _EJ0, we just call _SUN. This is also
similar to what we do for CPU hotplug.
Once we do this, there is no need to keep a separate SSDT for PCI hotplug.
Everything can reside in the same table.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
projects / seabios.git / log