First issue fixed:
For multi-socket CPU the current implementation emitted
Processor objects for cores in the first CPU only. This
commit fixes the bug by really emitting one Processor
object for each core. However, the unlikely case of mixed
CPU models is still not handled correctly.
Second issue fixed:
One loop was wrong in case a processor in the table declares
no P-states at all. The rewritten loop is safe. Some possibly
dangerous array lengths were also fixed.
Third issue: on MP-boards the recommended ramp-voltage (RVO) is 0mV
according to the BKDG. The current implementation always set it
to 25mV. This commit selects 0 or 25mV depending on CONFIG_MAX_PHYSICAL_CPUS.
Fourth issue: If a processor without PowerNow! support was inserted in a
system with coreboot configured with SET_FIDVID then the boot process hanged
mysteriously and very early. Apparently because init_fidvid_ap tampers with
non-existing registers. This commit fixes the bug by bailing out
from init_fidvid_ap if PowerNow! capability is missing.
Signed-off-by: Oskar Enoksson <enok@lysator.liu.se>
Change-Id: I61f6e2210b84ccba33a36c5efc866447b7134417
Reviewed-on: http://review.coreboot.org/239
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marcj303@gmail.com>
According to Rudolf Marek putting a memory instruction between
the CR0 write and the jmp in protected mode switching might hang the
machine. Move it after the jmp.
There might be a better solution for this, such as enabling the cache, as
keeping it disabled does not prevent cache poisoning attacks, so there is no
real point.
However, Intel docs say that SMM code in ASEG is always running uncached, so
we might want to consider running SMM out of TSEG instead, as well.
Signed-off-by: Stefan Reinauer <reinauer@google.com>
Change-Id: Id396acf3c8a79a9f1abcc557af6e0cce099955ec
Reviewed-on: http://review.coreboot.org/283
Reviewed-by: Sven Schnelle <svens@stackframe.org>
Tested-by: build bot (Jenkins)
It's only used in cbfs_and_run.c
Change-Id: Ibcfcefbeb0c5722eb3888f0d60127229a2badcf6
Signed-off-by: Stefan Reinauer <reinauer@google.com>
Reviewed-on: http://review.coreboot.org/273
Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
Tested-by: build bot (Jenkins)
If all console types are disabled, coreboot will fail to compile because
static code is unused. This patch fixes the issue.
Signed-off-by: Stefan Reinauer <reinauer@google.com>
Change-Id: Ie9c8bf2a78e3aeba4c2908b06bc03f0f5af37db2
Reviewed-on: http://review.coreboot.org/260
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marcj303@gmail.com>
Load an IDT with NULL limit to prevent the 16bit IDT being used
in protected mode before c_start.S sets up a 32bit IDT when entering
ram stage.
Signed-off-by: Stefan Reinauer <reinauer@google.com>
Change-Id: I8d048c894c863ac4971fcef8f065be6b899e1d3e
Reviewed-on: http://review.coreboot.org/259
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marcj303@gmail.com>
gcc 4.6 complains about unused but set variables in x86emu.
Particularly some variables are always set but only used in
debug mode, or when FPU support is enabled.
Change-Id: Ic53bd2303171ab717eb2d2c0ed72744d3eb6989e
Signed-off-by: Stefan Reinauer <reinauer@google.com>
Reviewed-on: http://review.coreboot.org/258
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marcj303@gmail.com>
- Intel option roms want an initialized i8259 or they will
throw an exception 6. This should be done in the southbridge
code, but that is executed much later than the VGA init, so
initialize the i8259 in src/devices/oprom/x86.c.
In the long run this will allow getting rid of some of the
ugly hacks in some AMD boards' romstage.c
- Don't overwrite the mode when copying mode info information back
from 0x600.
Change-Id: Idb01f13dbcd736d8d830b222ffe1ea85799fcd9c
Signed-off-by: Stefan Reinauer <reinauer@google.com>
Reviewed-on: http://review.coreboot.org/257
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marcj303@gmail.com>
- adds possibility to set a vesa mode without showing a bootsplash
- make bootsplash / mode setting code available in real mode.
Change-Id: I0045c9d75757657f4ce531889593102ea1e39ce5
Signed-off-by: Stefan Reinauer <reinauer@google.com>
Reviewed-on: http://review.coreboot.org/256
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marcj303@gmail.com>
- move int15 handler out of the generic code into the mainboard directories
of those mainboards that actually use it.
- move vbe headers to vbe.h
- move function prototypes used in native oprom code to x86.h
Change-Id: Idfff5e804ea328f7b5feebac72497c97329320ee
Signed-off-by: Stefan Reinauer <reinauer@google.com>
Reviewed-on: http://review.coreboot.org/255
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marcj303@gmail.com>
1.6.3 has a lot of benefits over the previous version, the two
most important being:
- working AHCI support
- compiles with gcc 4.6.x
Change-Id: Ie3a4d8f2624e0aa85e48ca09da53474c085838db
Signed-off-by: Stefan Reinauer <reinauer@google.com>
Reviewed-on: http://review.coreboot.org/253
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marcj303@gmail.com>
newer gcc versions generate ".section .text" instead of just ".text"
in their assembler output. This patch makes sure that we don't end up
with a superfluous ".section" that makes the build fail.
Add -Wno-unused-but-set-variable to CFLAGS if the flag exists.
Change-Id: I7f24c987433cc5886dde2af27498d3331cbda303
Signed-off-by: Stefan Reinauer <reinauer@google.com>
Reviewed-on: http://review.coreboot.org/252
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
That function broke SMP on Linux 2.4, now it works.
Change-Id: I4ddd25fef57bed64877959ca96cca68170042bca
Signed-off-by: Patrick Georgi <patrick.georgi@secunet.com>
Reviewed-on: http://review.coreboot.org/243
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marcj303@gmail.com>
Each variable is essentially unused or incorrect.
Change-Id: I4d2a10c9b45306ac6e6026a31765d3b912fd855c
Signed-off-by: Patrick Georgi <patrick.georgi@secunet.com>
Reviewed-on: http://review.coreboot.org/242
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marcj303@gmail.com>
Otherwise with a coreboot log on COM2 (which doesn't work) the boot
process takes eons.
Change-Id: I886f98b715c1f384c8693f2977671ff15897b5a5
Signed-off-by: Patrick Georgi <patrick.georgi@secunet.com>
Reviewed-on: http://review.coreboot.org/241
Reviewed-by: Marc Jones <marcj303@gmail.com>
Tested-by: build bot (Jenkins)
The last couple of lines of every mptable function were mostly
identical. Refactor into common code, a new function mptable_finalize.
Coccinelle script:
@@
identifier mc;
@@
(
-mc->mpe_checksum = smp_compute_checksum(smp_next_mpc_entry(mc), mc->mpe_length);
-mc->mpc_checksum = smp_compute_checksum(mc, mc->mpc_length);
-printk(BIOS_DEBUG, "Wrote the mp table end at: %p - %p\n", mc, smp_next_mpe_entry(mc));
-return smp_next_mpe_entry(mc);
+return mptable_finalize(mc);
|
-mc->mpe_checksum = smp_compute_checksum(smp_next_mpc_entry(mc), mc->mpe_length);
-mc->mpc_checksum = smp_compute_checksum(mc, mc->mpc_length);
-return smp_next_mpe_entry(mc);
+return mptable_finalize(mc);
)
Change-Id: Ib2270d800bdd486c5eb49b328544d36bd2298c9e
Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Reviewed-on: http://review.coreboot.org/246
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marcj303@gmail.com>
We copied pretty much the same code for generating mptable entries for
local interrupts (with some notable exceptions).
This change moves these lines into a generic function "mptable_lintsrc"
and makes use of it in many places.
The remaining uses of smp_write_lintsrc should be reviewed and replaced
by mptable_lintsrc calls where possible, and smp_write_lintsrc made static.
This patch was generated using Coccinelle:
@@
expression mc;
expression isa_bus;
@@
-smp_write_lintsrc(mc, mp_ExtINT, MP_IRQ_TRIGGER_DEFAULT|MP_IRQ_POLARITY_DEFAULT, isa_bus, 0x0, MP_APIC_ALL, 0x0);
-smp_write_lintsrc(mc, mp_NMI, MP_IRQ_TRIGGER_DEFAULT|MP_IRQ_POLARITY_DEFAULT, isa_bus, 0x0, MP_APIC_ALL, 0x1);
+mptable_lintsrc(mc, isa_bus);
@@
expression mc;
expression isa_bus;
@@
-smp_write_lintsrc(mc, mp_ExtINT, MP_IRQ_TRIGGER_EDGE|MP_IRQ_POLARITY_HIGH, isa_bus, 0x0, MP_APIC_ALL, 0x0);
-smp_write_lintsrc(mc, mp_NMI, MP_IRQ_TRIGGER_EDGE|MP_IRQ_POLARITY_HIGH, isa_bus, 0x0, MP_APIC_ALL, 0x1);
+mptable_lintsrc(mc, isa_bus);
@m@
identifier mc;
expression BUS;
@@
-#define IO_LOCAL_INT(type, intr, apicid, pin) smp_write_lintsrc(mc, (type), MP_IRQ_TRIGGER_EDGE | MP_IRQ_POLARITY_HIGH, BUS, (intr), (apicid), (pin));
...
-IO_LOCAL_INT(mp_ExtINT, 0x0, MP_APIC_ALL, 0x0);
-IO_LOCAL_INT(mp_NMI, 0x0, MP_APIC_ALL, 0x1);
+mptable_lintsrc(mc, BUS);
Change-Id: I97421f820cd039f5fd753cb0da5c1cca68819bb4
Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Reviewed-on: http://review.coreboot.org/244
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marcj303@gmail.com>
Use the SSDT autogen infrastructure to support the automatic reserved resources,
automatic P-state generation and automatic _CRS PCI0 method.
Change-Id: Ic56a92eeb70a0a2a2d6de2507009ec3a832c83b3
Signed-off-by: Rudolf Marek <r.marek@assembler.cz>
Reviewed-on: http://review.coreboot.org/251
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
Add a ACPI Source Language snippet to superio/w83627hf which maps the
SuperIO and most of the logical devices to PnP devices, exposing
configuration options and chip power management to the OS.
Written using the Winbond W83627HF/F datasheet.
Change-Id: I1108d29b341ef78fe7f1e574f98b680aada39daf
Signed-off-by: Christoph Grenz <christophg+cb@grenz-bonn.de>
Reviewed-on: http://review.coreboot.org/223
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marcj303@gmail.com>
Add a ACPI Source Language snippet which if included as
shown in the comments in the file, exposes the 4 possible
temperature sensors in the CPU as ACPI thermal zones.
Change-Id: I94dd773108e348a0fdb9d2f8d6cfe415d5fa0339
Signed-off-by: Christoph Grenz <christophg+cb@grenz-bonn.de>
Reviewed-on: http://review.coreboot.org/222
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marcj303@gmail.com>
SB800 RAID ROM require to put the misc ROM to specific position,
this patch enable user to put the RAID misc ROM to the right place
in the coreboot image.
Change-Id: I4fc64df8e091fb0cccd063826ab31a4f198942d1
Signed-off-by: Kerry She <kerry.she@amd.com>
Signed-off-by: Kerry She <shekairui@gmail.com>
Reviewed-on: http://review.coreboot.org/249
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marcj303@gmail.com>
In order to make sure AHCI/RAID ROM works correctly
For SB800_SATA_AHCI or SB800_SATA_RAID mode, SATA should
enable bus master and the ahci also should be enabled.
Change-Id: I9d9c557816d364d8373fe343860ad5fe45988200
Signed-off-by: Kerry She <kerry.she@amd.com>
Signed-off-by: Kerry She <shekairui@gmail.com>
Reviewed-on: http://review.coreboot.org/248
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marcj303@gmail.com>
Add this option to enable/disable SATA IDE Combined Mode feature
Change-Id: I1ab8acd27947a71baf954f44d0741f81f48e5541
Signed-off-by: Kerry Sheh <kerry.she@amd.com>
Signed-off-by: Kerry Sheh <shekairui@gmail.com>
Reviewed-on: http://review.coreboot.org/231
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marcj303@gmail.com>
sb800 cimx entry sb_Before_Pci_Init was called in the device 16.2
enable_dev() function. If the devicetree don't have this device,
then sb_Before_Pci_Init will not get called.
So the missing sb800 USB3 devicees was add to the mainboard devicetree.
Because of no physical usb connector connected to USB3, the USB3 device setting was off.
Change-Id: If060ccb43df7fbe88bafc61e9e600a9120575437
Signed-off-by: Kerry Sheh <kerry.she@amd.com>
Signed-off-by: Kerry Sheh <shekairui@gmail.com>
Reviewed-on: http://review.coreboot.org/232
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marcj303@gmail.com>
Doing a call before the payload has set up its stack is risky. The stack may
not be in a favorable location. Normally this is not an issue with coreboot
or other well behaved callers.
Change-Id: Ie6f6748a471324b29ebad045c807dfc9f4b92034
Signed-off-by: Marc Jones <marcj303@gmail.com>
Reviewed-on: http://review.coreboot.org/240
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
AMD K8 rev F and later implements a bit SYSCFG_MSR_TOM2WB to
mark dram memory above 4GB as WB. However, AMD K8
rev E and earlier don't implement this bit and therefore need
MTRR spanning dram memory above 4GB. The current implementation
of amd_setup_mtrrs never generate MTRR above 4GB.
This caused memory > 4GB not to be recognized in e.g. Linux on those
rev E or older platforms. This commit should fix that bug.
Signed-off-by: Oskar Enoksson <enok@lysator.liu.se>
Change-Id: Ie568a52a8eb355969c86964d5afc4692e60f69c1
Reviewed-on: http://review.coreboot.org/238
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marcj303@gmail.com>
sb800 cimx entry sb_Before_Pci_Init was called in the device 16.2
enable_dev() function. If the devicetree don't have this device,
then sb_Before_Pci_Init will not get called.
Change-Id: I76ebad842e90b0f740abbec031165d7c39a80abf
Signed-off-by: Kerry Sheh <kerry.she@amd.com>
Signed-off-by: Kerry Sheh <shekairui@gmail.com>
Reviewed-on: http://review.coreboot.org/230
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marcj303@gmail.com>
Using micro CIMX_OPTION_ENABLED/CIMX_OPTION_DISABLED to
configure SataIdeCombinedMode is wrong.
sbPowerOnInit() use SataIdeCombinedMode to determine whether hide the IDE controller
0: IDE controller is exposed and Combined Mode is enabled.
SATA controller has control over Port0 through Port3,
IDE controller has control over Port4 and Port5
1: IDE controller is hidden and Combined Mode is disabled,
SATA controller has full control of all 6 Ports when operating in non-IDE mode
Change-Id: I32e7101737f1dbfff49daa58670e6820b476b250
Signed-off-by: Kerry Sheh <kerry.she@amd.com>
Signed-off-by: Kerry Sheh <shekairui@gmail.com>
Reviewed-on: http://review.coreboot.org/229
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marcj303@gmail.com>
If sb800 sata was configured as ahci or raid mode,
give the option to add ROM files.
Change-Id: I87a7814930ce3a7c38cde1e235d151223eea2107
Signed-off-by: Kerry Sheh <kerry.she@amd.com>
Signed-off-by: Kerry Sheh <shekairui@gmail.com>
Reviewed-on: http://review.coreboot.org/225
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marcj303@gmail.com>
sb800 SATA device have different device id with different configure
mode, 4392h for RAID mode, 4393h for RAID5 mode
Change-Id: If54f7751f531c94ee725309a2a5c255390935ead
Signed-off-by: Kerry Sheh <kerry.she@amd.com>
Signed-off-by: Kerry Sheh <shekairui@gmail.com>
Reviewed-on: http://review.coreboot.org/226
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marcj303@gmail.com>
The hp/dl145_g1 motherboard did not work since commit
1f7d3c5672 (svn 6124). That commit added
TINY_BOOTBLOCK for amd8111 southbridge. The result was that the boot process
stopped very early (no console output whatsoever). The same symptom was
reported on other AMDK8 based boards with amd8111 southbridge chips. This
commit seems to fix the bug. It adds a bootblock.c under
src/northbridge/amd/amdk8 that calls enumerate_ht_chains. Probably the
problem was that enum_ht_chains needs to be called before the southbridge
bootblock.c function, not after.
Change-Id: I74fb892aa39048e2d0e76c081b713f825d67f2d4
Signed-off-by: Oskar Enoksson <enok@lysator.liu.se>
Reviewed-on: http://review.coreboot.org/235
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
Reviewed-by: Marc Jones <marcj303@gmail.com>
the patch file comes from
src/vendorcode/amd/agesa/f12/Proc/CPU/Family/0x10/RevE
/F10MicrocodePatch010000bf.c
Change-Id: If701c8a908edf1c486665d3ce4df65da0f65c802
Signed-off-by: QingPei Wang <wangqingpei@gmail.com>
Reviewed-on: http://review.coreboot.org/202
Tested-by: build bot (Jenkins)
Reviewed-by: Peter Stuge <peter@stuge.se>
Gcc 4.1 comes with an SSP https://wiki.ubuntu.com/GccSsp
This is disabled to work around '__stack_chk_fail' symbol not found failures
http://www.coreboot.org/FAQ/Obsolete#How_do_I_fix_stack_chk_fail_errors.3F
The presence of -fno-stack-protector is tested for automatically by configure.
Change-Id: I28ef158829f5935f985cfd5a5440733685cf479a
Reported-by: Raymond Danks <raymonddanks@gmail.com>
Signed-off-by: Raymond Danks <raymonddanks@gmail.com>
Reviewed-on: http://review.coreboot.org/112
Tested-by: build bot (Jenkins)
Reviewed-by: Peter Stuge <peter@stuge.se>
LXBIOS and LXB-DSDT are not used in other parts of the tree.
Make names consistent across the tree.
Signed-off-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
Change-Id: I91caeac09fd2401a36e53bd061d249b236a48e43
Reviewed-on: http://review.coreboot.org/224
Tested-by: build bot (Jenkins)
Reviewed-by: Peter Stuge <peter@stuge.se>
Change-Id: Ic960fe09fbed2c8a31c7c9ac2c54f6c88efebed3
Signed-off-by: Marc Jones <marcj303@gmail.com>
Reviewed-on: http://review.coreboot.org/219
Reviewed-by: Frank Vibrans III <frank.vibrans@amd.com>
Tested-by: build bot (Jenkins)
Change-Id: Ibd87422680350c112eabe1bb73b237031c3e9d6b
Signed-off-by: Marc Jones <marcj303@gmail.com>
Reviewed-on: http://review.coreboot.org/220
Tested-by: build bot (Jenkins)
Reviewed-by: Frank Vibrans III <frank.vibrans@amd.com>
These are the changes for the AMD Persimmon mainboard
required to support the update of the AMD Family 14
cpu to rev C0. There are many warning fixes; the agesa-
wrapper.c file has been changed to fix the amdinitlate
and amdlaterunaptask routines, and more.
Change-Id: I6de43379a2819cea5169db5f21d4841f9a4942a7
Signed-off-by: Frank Vibrans <frank.vibrans@amd.com>
Signed-off-by: efdesign98 <efdesign98@gmail.com>
Reviewed-on: http://review.coreboot.org/137
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marcj303@gmail.com>
Oskar Enoksson