MASK_STATUS disables interrupt status generation for the entire GPIO
controller when any debounce register is configured. This causes
problems when the kernel is loading drivers because we could lose
interrupts for previously loaded devices.
sb_program_gpios is also not setup to wait when configuring
PAD_DEBOUNCE, so there is a potential that we could lose the interrupt
status enable bit for other registers. By disabling MASK_STATUS we avoid
that problem.
BUG=b:113880780
BRANCH=none
TEST=Ran a reboot stress test that concluded that we are no longer
losing TPM interrupts while booting.
Change-Id: Ife1db3b1449f205092509595cbc3eca511bff57a
Signed-off-by: Raul E Rangel <rrangel@chromium.org>
Reviewed-on: https://review.coreboot.org/c/31082
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Daniel Kurtz <djkurtz@google.com>
It is possible to configure debounce, but leave it disabled by specifying
a 0 value for the timeout. Add a define for allowing to do so via the
PAD_DEBOUNCE() macro.
Signed-off-by: Daniel Kurtz <djkurtz@chromium.org>
BUG=b:113880780
BRANCH=none
TEST=compile
Change-Id: I9de61297b0677cc904535a51c16970eecb52021d
Reviewed-on: https://review.coreboot.org/c/30998
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Martin Roth <martinroth@google.com>
Reviewed-by: Raul Rangel <rrangel@chromium.org>
This reverts commit b82afce18a.
Reason for revert: This causes depthcharge to not boot due to TPM timeout errors. Because there is no wait after setting the debounce register, we lose data because the read-modify-write loses the interrupt status bit.
e.g., GPIO 5 sets debounce, without a wait. Then GPIO 9 has it's debounce set. Because the interrupt controller is masking the interrupt enable status bit, the read-modify-write for GPIO9 loses the interrupt enable status bit and it never gets set again. This causes the interrupt to never latch.
We should possibly make depthcharge set the interrupt enable status bit for latched GPIOs.
Change-Id: Idd7259b14b24c441529d64e173be9faec03f4fc8
Signed-off-by: Raul E Rangel <rrangel@chromium.org>
Signed-off-by: Martin Roth <martinroth@google.com>
Reviewed-on: https://review.coreboot.org/c/30981
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Daniel Kurtz <djkurtz@google.com>
FT4 has a strange property where whenever the debounce registers for any
one gpio are changed, the FT4 disables interrupt propagation for ALL
gpio irqs for ~4ms.
In other words, if an edge interrupt of one gpio happens exactly during
this debounce-irq-off window immediately following the configuration of
another gpio, the interrupt will be lost.
It is quite difficult to deal with this in the kernel, since during kernel
boot time, drivers & devices are probed asynchronously, meaning it may
happen that an already loaded driver may miss an interrupt when some
later driver is being probed and configuring its gpio interrupt.
To eliminate this possibility, we pre-configure the debounce registers in
ram stage for all gpios that will be used as irqs later by the kernel
using the same configuration as used by the kernel, as per this table:
IRQ Debounce
Edge Remove Glitch
Level High Preserve Low Glitch
Level Low Preserve High Glitch
Signed-off-by: Daniel Kurtz <djkurtz@chromium.org>
BUG=b:113880780
BRANCH=none
TEST=Reboot stress test grunt (>100 times); no messages in dmesg like:
tpm tpm0: Timeout waiting for TPM ready
Change-Id: I94c7ecfb14e5bb209b3598e10287c80eb19da25b
Reviewed-on: https://review.coreboot.org/c/30921
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Martin Roth <martinroth@google.com>
Invert the default instead of selecting it everywhere. Restores the
ability to use its Kconfig prompt.
Beside Qemu targets, the only platforms that didn't select it seem
to be samsung/exynos5420, intel/cannonlake, and intel/icelake. The
latter two were about to be patched anyway.
Change-Id: I7c5b671b7dddb5c6535c97c2cbb5f5053909dc64
Signed-off-by: Nico Huber <nico.h@gmx.de>
Reviewed-on: https://review.coreboot.org/c/30891
Reviewed-by: Lijian Zhao <lijian.zhao@intel.com>
Reviewed-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Currently this needlessly initializes the hardware in the both the
romstage and the bootblock, but it works.
Build option is renamed to USBDEBUG_IN_PRE_RAM to reflect the
use better, related support files can be built to pre-ram stages
regardless of usbdebug being enabled or not.
Tested on Google/peppy (adapted to C_ENVIRONMENT_BOOTBLOCK).
Change-Id: Ib77f2fc7f3d8fa524405601bae15cce9f76ffc6f
Signed-off-by: Arthur Heymans <arthur@aheymans.xyz>
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/c/30480
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
disable_all_smi_status() was not clearing SMI_EVENT_STATUS. This caused
us to complain in the eventlog (ELOG_SLEEP_PENDING_GPE0_WAKE) and then
wake early from sleep when waiting for a cr50 reset to turn on a cr50
update.
BUG=b:121203745
TEST=Careena remains in S5 until cr50 reset after cr50 update, and
ELOG_SLEEP_PENDING_GPE0_WAKE is no longer seen in eventlog.
Change-Id: I2eec014109249d5c3574c4dbdec5569e2a0bfc8e
Signed-off-by: Edward Hill <ecgh@chromium.org>
Reviewed-on: https://review.coreboot.org/c/30304
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-by: Raul Rangel <rrangel@chromium.org>
Reviewed-by: Martin Roth <martinroth@google.com>
This is needed so the next patch can set up GPIOs before
AGESA runs.
BUG=b:120436919
TEST=Verified romstage mainboard code runs before AGESA
Change-Id: I76c035e166cd64382b52dff5ae00a6f115cbac9b
Signed-off-by: Martin Roth <martinroth@chromium.org>
Reviewed-on: https://review.coreboot.org/c/30038
Reviewed-by: Daniel Kurtz <djkurtz@google.com>
Reviewed-by: Marshall Dawson <marshalldawson3rd@gmail.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
AMD traditionally claims the resource at I/O port 61 for the onboard
PC-AT speaker. In later designs, the speaker may be omitted in favor
of routing the SPKR signal to the codec.
Some systems implement neither, and for those it is not correct to
identify the resource as a speaker. Modify the EISAID reported to
the OS depending on the system design. The default is that port 61
is reported as reserved. In order to report a speaker, add #define
in mainboard//dsdt.asl.
TEST=check /proc/ioports and iasl -d for both ways using a Grunt
BUG=b:117818432
Change-Id: I33aafb187f9fea7b38aae43c399292c7521fcfc4
Signed-off-by: Marshall Dawson <marshalldawson3rd@gmail.com>
Reviewed-on: https://review.coreboot.org/c/30037
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Martin Roth <martinroth@google.com>
When building grunt with flags set to detect variables that get a value but
then are unused, there are 5 instances that causes error (unused variable).
In most cases it's enough to simply remove the variable. Other instances,
is better to simply use the variables (one instance it's a return value, on
the other instance using the variables makes code more readable).
BUG=b:120260448
TEST=Build and boot grunt.
Change-Id: I0d00fb6a42db20afb34c76b9445a741a57096ead
Signed-off-by: Richard Spiegel <richard.spiegel@silverbackltd.com>
Reviewed-on: https://review.coreboot.org/c/29985
Reviewed-by: HAOUAS Elyes <ehaouas@noos.fr>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Instead of ELOG_GSMI_APM_CNT use APM_CNT_ELOG_GSMI and define it in
cpu/x86/smm.h
Change-Id: I3a3e2f823c91b475d1e15b8c20e9cf5f3fd9de83
Signed-off-by: Patrick Georgi <pgeorgi@google.com>
Reviewed-on: https://review.coreboot.org/c/30022
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Replace IO access to ACPI registers with the new MMIO access functions.
BUG=b:118049037
TEST=Build and boot grunt. Test ACPI related functionality.
Change-Id: I7544169bb21982fcf7b1c07ab7c19c6f5e65ad56
Signed-off-by: Richard Spiegel <richard.spiegel@silverbackltd.com>
Reviewed-on: https://review.coreboot.org/c/29296
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Marshall Dawson <marshalldawson3rd@gmail.com>
ACPI registers can be accessed through IO or through MMIO. However, the
offset relationship is not one to one. Therefore, definitions with the
correct offset for MMIO access are needed.
BUG=b:118049037
TEST=Use Chrome OS IOTOOLS io_readxx and mem_readxx to find the correct
relationship between ACPI IO and ACPI MMIO.
Change-Id: Id20754c0fc0af35bc9eb1a4b40c62fbf9ed6304d
Signed-off-by: Richard Spiegel <richard.spiegel@silverbackltd.com>
Reviewed-on: https://review.coreboot.org/c/29294
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Marshall Dawson <marshalldawson3rd@gmail.com>
Now that the relationship between IO access and MMIO access has been
established, create read/write functions to access ACPI standard registers
through MMIO.
BUG=b:118049037
TEST=Build grunt
Change-Id: I32c26f342885c0d99b082be98730edcf16ab5dfc
Signed-off-by: Richard Spiegel <richard.spiegel@silverbackltd.com>
Reviewed-on: https://review.coreboot.org/c/29295
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Marshall Dawson <marshalldawson3rd@gmail.com>
Makes vboot measured boot mode available for all boards.
* Increase Tegra210 and Rockchip3228 SRAM for
romstage/verstage.
* Add missing files for Intel apollolake and
AMD stoneyridge as TPM driver target.
Change-Id: I35a85b8f137f28cd9960f2c5ce95f8fa31185b82
Signed-off-by: Philipp Deppenwiese <zaolin.daisuki@gmail.com>
Reviewed-on: https://review.coreboot.org/c/29840
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Werner Zeh <werner.zeh@siemens.com>
Reviewed-by: Julius Werner <jwerner@chromium.org>
Some "magic" numbers became public available registers/bits after the code
was originally written. Find all magic numbers, and if available in a public
BKDG than replace them with literals.
BUG=b:117648026
TEST=Build and boot grunt.
Change-Id: I96ac59fd92c4a5e27c3836f77bf6633e9b0c4990
Signed-off-by: Richard Spiegel <richard.spiegel@silverbackltd.com>
Reviewed-on: https://review.coreboot.org/c/29198
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
There are several files under soc/amd that use units defined by file
porting.h. These units use upper case, and are not recognized by checkpatch,
thus causing problems when defining a pointer (request to use space before
and after the star symbol). These are the definitions from porting.h showing
the units that this patch will change and their coreboot definitions (not all
are actually used):
typedef uintptr_t UINTN;
typedef int64_t INT64;
typedef uint64_t UINT64;
typedef int32_t INT32;
typedef uint32_t UINT32;
typedef int16_t INT16;
typedef uint16_t UINT16;
typedef int8_t INT8;
typedef uint8_t UINT8;
typedef char CHAR8;
typedef unsigned short CHAR16;
typedef unsigned char BOOLEAN;
typedef void VOID;
BUG=b:118775313
TEST=Build and boot grunt.
Change-Id: Ic1bd64d6224a030a65d23decabf0e602cee02871
Signed-off-by: Richard Spiegel <richard.spiegel@silverbackltd.com>
Reviewed-on: https://review.coreboot.org/29520
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Marshall Dawson <marshalldawson3rd@gmail.com>
Function write_pci_cfg_irqs() has "no function" variables. One variable is
set and never used, the other is only used to control a print. Remove them.
BUG=b:117950052
TEST=Build grunt.
Change-Id: Icd98db3e794e609b112f15979a3a00a2977a0fdb
Signed-off-by: Richard Spiegel <richard.spiegel@silverbackltd.com>
Reviewed-on: https://review.coreboot.org/29244
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Marshall Dawson <marshalldawson3rd@gmail.com>
1. Add two parameters for panel initialization timing.
> lvds_poseq_varybl_to_blon
> lvds_poseq_blon_to_varybl
2. The BL_PWM is controlled by APU_EDP_BKLTEN_L/APU_DP_VARY_BL/
EDP_BKLTEN_L, so move APU_EDP_BKLTEN_L to early init stage,
and be enabled depends on SKU, thus we can control the delay
time by config APU_DP_VARY_BL.
BUG=b:118011567
TEST=emerge-grunt coreboot.
Change-Id: Ib20c48813b208d697b950b2f02a70a690e483fdb
Signed-off-by: Chris Wang <chris.wang@amd.corp-partner.google.com>
Reviewed-on: https://review.coreboot.org/29469
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Martin Roth <martinroth@google.com>
There some files that do have at least 1 line over the 80 characters limit.
Find and fix them.
BUG=b:117950052
TEST=Build grunt.
Change-Id: I1083a7559919e05a3e3a2dac99f571c161bb4c27
Signed-off-by: Richard Spiegel <richard.spiegel@silverbackltd.com>
Reviewed-on: https://review.coreboot.org/29228
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Marshall Dawson <marshalldawson3rd@gmail.com>
There are functions within heapmanager.c that have a space between the
function name and open parenthesis. Remove these spaces.
BUG=b:117950051
TEST=build grunt.
Change-Id: I2120d9d5f663453b6201d1872f29c6dc4abd6191
Signed-off-by: Richard Spiegel <richard.spiegel@silverbackltd.com>
Reviewed-on: https://review.coreboot.org/29230
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Marshall Dawson <marshalldawson3rd@gmail.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Pointer math with void pointers is illegal in many compilers, though it
works with GCC because it assumes size of void to be 1. Change the pointers
or add parenthesis to force a proper order that will not cause compile
errors if compiled with a different compiler, and more importantly, don't
have unsuspected side effects.
BUG=b:118484178
TEST=Build and boot grunt.
Change-Id: Ibfeb83893f09cb897d459856aff2a4ab2a74e6e5
Signed-off-by: Richard Spiegel <richard.spiegel@silverbackltd.com>
Reviewed-on: https://review.coreboot.org/29337
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Marshall Dawson <marshalldawson3rd@gmail.com>
In commit 41baf0c3ff (soc/amd/stoneyridge: Remove dev_find_slot where
possible), the register being read was changed accidentally from
HT_DEV (Device 18h, Func 0) to NB_DEV (Device 18h, Func 5)
This doesn't return the correct value, and causes Grunt to reboot.
BUG=b:118721473
TEST=Boot grunt
Change-Id: I7b73358a074dd27639aafead7c8b39f0fad5685f
Signed-off-by: Martin Roth <martinroth@google.com>
Reviewed-on: https://review.coreboot.org/29367
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Daniel Kurtz <djkurtz@google.com>
Reviewed-by: Richard Spiegel <richard.spiegel@silverbackltd.com>
Reviewed-by: Raul Rangel <rrangel@chromium.org>
The function domain_read_resources() didn't have any code to actually
reserve any resources - it was just creating an empty resource entry.
I looked at fixing it to actually reserve the space, but the values in
the registers at the point when this runs aren't the final values that
we want to reserve anyway, they're temp values with a range much larger
than we want to reserve.
I next looked at moving the amd_initcpuio() function earlier so that we
could get the correct values for the registers, but even that doesn't
give us what we really want.
Ultimately removing this whole function seems to be the right thing.
BUG=None
TEST=Verify that the only resource that changes is the empty resource:
PCI: 00:18.0 resource base 0 size 0 align 0 gran 0 limit 0 flags 1 index 1080
Change-Id: I83bd3ea8db141416632c12fc883386070363f2f1
Signed-off-by: Martin Roth <martinroth@google.com>
Reviewed-on: https://review.coreboot.org/29345
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Marshall Dawson <marshalldawson3rd@gmail.com>
Reviewed-by: Richard Spiegel <richard.spiegel@silverbackltd.com>
Instead of forcing the IOMMU to be enabled, change it to only be enabled
if the device is enabled in devicetree.
BUG=b:118612241
TEST=Verify that IOMMU is disabled.
Change-Id: I6cfd6c81f47de23c54a49ec7cf87b219215ced5e
Signed-off-by: Martin Roth <martinroth@google.com>
Reviewed-on: https://review.coreboot.org/29343
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Raul Rangel <rrangel@chromium.org>
Reviewed-by: Daniel Kurtz <djkurtz@google.com>
The procedure dev_find_slot has 3 main uses. To find configuration
(devicetree), to verify if a particular device is enabled at build \
time, and to get the address for PCI access while in bootblock/romstage.
The third use can be hidden by using macros defined in pci_devs.h,
making it very clear what PCI device is being accessed. replace the
temporary pointers to device used with PCI access with SOC_XXX_DEV where
XXX is the device being accessed, and remove the setting of the temporary
pointers.
BUG=b:117917136
TEST=Build grunt.
Change-Id: Ic38ea04bfcc1ccaa12937b19e9442a26d869ef11
Signed-off-by: Richard Spiegel <richard.spiegel@silverbackltd.com>
Reviewed-on: https://review.coreboot.org/29227
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Marshall Dawson <marshalldawson3rd@gmail.com>
In function agesa_GfxGetVbiosImage(), the function name is used in a print
string. Use __func__ instead.
BUG=b:117642170
TEST=Build grunt.
Change-Id: I95a042bd95cc729305a8a008e3bb464f60c2668d
Signed-off-by: Richard Spiegel <richard.spiegel@silverbackltd.com>
Reviewed-on: https://review.coreboot.org/29246
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Marshall Dawson <marshalldawson3rd@gmail.com>
File ramtop.c has one instance of if()/else where the if tests for top mem
in lower 4GiB, and returns just before the "else" statement. These "else"
statements are not needed.
BUG=b:117648025
TEST=Build and boot grunt.
Change-Id: Iba16a416e78dae75a95a11d38179161c5a11b2ad
Signed-off-by: Richard Spiegel <richard.spiegel@silverbackltd.com>
Reviewed-on: https://review.coreboot.org/29247
Reviewed-by: Marshall Dawson <marshalldawson3rd@gmail.com>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Accurately reflect the intention of the syslimit value returned
from AmdInitPost(). Assume FFs for the non-present bits.
BUG=b:118178425
TEST=Boot Grunt and verify reported value = TOM2-1.
Change-Id: Ie8ea4fcbfd52c46ad441890f0decaf0f55816cfd
Signed-off-by: Marshall Dawson <marshalldawson3rd@gmail.com>
Reviewed-on: https://review.coreboot.org/29229
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Martin Roth <martinroth@google.com>
The file smbus.asl has 0 bytes (no content). Now that it's no longer included,
remove it.
BUG=b:117814641
TEST=Build grunt and gardenia.
Change-Id: I66389c721e272053d86357f71a6d1242ca767edd
Signed-off-by: Richard Spiegel <richard.spiegel@silverbackltd.com>
Reviewed-on: https://review.coreboot.org/29182
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Martin Roth <martinroth@google.com>
Add an `amdblocks` internal API and rename
soft_reset() => warm_reset()
hard_reset() => cold_reset()
as these terms are commonly used in the surrounding code.
On Stoney Ridge, make board_reset() call cold_reset() to keep
current behaviour of common code calling hard_reset(). But add
a TODO if this is intended.
Note: Stoney Ridge is using CF9 for the actual reset but the
configuration for a cold reset doesn't use the usual full reset
bit but some other mechanism.
Change-Id: Id33eda676d79529db759b85fa8e28386846e6fa4
Signed-off-by: Nico Huber <nico.h@gmx.de>
Reviewed-on: https://review.coreboot.org/29053
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Richard Spiegel <richard.spiegel@silverbackltd.com>
Phase 1. Due to the size of the effort, this CL is broken into several
phases.
Change-Id: I0236e0960cd1e79558ea50c814e1de2830aa0550
Signed-off-by: Elyes HAOUAS <ehaouas@noos.fr>
Reviewed-on: https://review.coreboot.org/29065
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Richard Spiegel <richard.spiegel@silverbackltd.com>
Register 0x40 of miscellaneous MMIO is double defined, with different names,
which makes it confusing. Eliminate MISC_MISC_CLK_CNTL_1, and move its only
bit definition to MISC_CLK_CNTL1 (which is correctly placed among MMIO
registers.
BUG=b:117818431
TEST=Build grunt.
Change-Id: I5ca5045498b8a81943282e0d6ecfbaecbd600d19
Signed-off-by: Richard Spiegel <richard.spiegel@silverbackltd.com>
Reviewed-on: https://review.coreboot.org/29175
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Marshall Dawson <marshalldawson3rd@gmail.com>
