To ensure that the command1 write which sets the "go" bit completes before
other reads to the device. Otherwise, there's a race condition where those
register values might still have their values from the last transfer. With
different SPI clock frequencies, that could lead to spi_delay being told there
were negative bytes still to send. Its expected delay would wrap to a negative
value, that was passed to udelay, and the system would sit there for 4 seconds
not doing anything.
BUG=None
TEST=Built and booted on nyan. Set the SPI bus frequency to a value which was
causing the 4+ second delay and verified that it no longer happened.
BRANCH=None
Original-Change-Id: I8b4090efc69f34d0413e3f63c59c1825dd151cec
Original-Signed-off-by: Gabe Black <gabeblack@google.com>
Original-Reviewed-on: https://chromium-review.googlesource.com/193347
Original-Reviewed-by: Gabe Black <gabeblack@chromium.org>
Original-Commit-Queue: Gabe Black <gabeblack@chromium.org>
Original-Tested-by: Gabe Black <gabeblack@chromium.org>
(cherry picked from commit d7ea9febdf2c5942f81607ee6ded786c9a8954bb)
Signed-off-by: Marc Jones <marc.jones@se-eng.com>
Change-Id: I095bfc745eda37b8e666475ceb41684152f3709a
Reviewed-on: http://review.coreboot.org/7737
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
This fixes two problems with the clock configuration on tegra124. First, the
macro which set up the i2c clocks tried to account for the fact that the i2c
divisor's lsb represents 1.0 where it normally represents 0.5 by multiplying
the target frequency by 2. That doesn't work, unfortunately, because the
divisor is actually n + 1, and what n + 1 means depends on where the one's
place is in the divisor.
Also, when calculating the divisor, the standard C division operator uses
truncation to deal any remainder which tends to make the divisor smaller. That
has the effect of making the output frequency higher than what was requested.
Since it's usually safer to undershoot a frequency than overshoot it, this
change makes those divisions round up instead.
Finally, the hand tuned temporary UART clock configuration was adjusted so
that it still ends up with the same divisor. Without that, very early output
from the bootblock is garbled, specifically the coreboot welcome banner,
build timestamp, etc.
BUG=chrome-os-partner:27220
TEST=Built and booted on nyan. Used a logic analyzer to verify that the TPM
i2c bus ran at 400KHz instead of 660KHz, and that the divisor was the expected
value. Measured boot time with and without EFS and verified that there was no
change. Spot checked the output for errors and verified that none of the
bootblock output was garbled.
BRANCH=None
Had to add the stdlib.h from 89ed6c that hadn't been merged correctly.
Original-Change-Id: I7e948c361ed4bf58c608627d32f2e3424faea1fb
Original-Signed-off-by: Gabe Black <gabeblack@google.com>
Original-Reviewed-on: https://chromium-review.googlesource.com/193362
Original-Reviewed-by: Julius Werner <jwerner@chromium.org>
Original-Commit-Queue: Gabe Black <gabeblack@chromium.org>
Original-Tested-by: Gabe Black <gabeblack@chromium.org>
(cherry picked from commit 164f7010a47d3bbdbc8bb572106140ae186f3807)
Signed-off-by: Marc Jones <marc.jones@se-eng.com>
Change-Id: I317b66eda929c0e5a5832adca267b8b54c6aae34
Reviewed-on: http://review.coreboot.org/7736
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
To read EDID, we need to access I2C via DP AUX channel.
BRANCH=none
BUG=chrome-os-partner:25933
TEST=emerge-nyan coreboot chromeos-bootimage
Original-Change-Id: I2666b5d46843485b79265a537f19bd8eab5e1a26
Original-Signed-off-by: Hung-Te Lin <hungte@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/188858
Original-Reviewed-by: Gabe Black <gabeblack@chromium.org>
Original-Commit-Queue: Gabe Black <gabeblack@chromium.org>
(cherry picked from commit 8f8e98ff5038b57f89332aee75573095c3933dd2)
Signed-off-by: Marc Jones <marc.jones@se-eng.com>
Change-Id: I5b1b6ab2940c8265483059fd94a2c4db2a41144a
Reviewed-on: http://review.coreboot.org/7735
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
Tested-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
If EFS is enabled and vboot didn't tell us it's going to use the display, we
can skip initializing it and save some boot time.
BUG=chrome-os-partner:27094
TEST=Built and booted on nyan without EFS in recovery mode and normal mode.
Built and booted on nyan with EFS in recovery mode and normal mode. Verified
that in normal mode with EFS the display initialization was skipped and boot
time was essentially the same as when display initialization was simply
commented out.
BRANCH=None
Original-Change-Id: I1e2842b57a38061f40514407c8fab1e38b75be80
Original-Signed-off-by: Gabe Black <gabeblack@google.com>
Original-Reviewed-on: https://chromium-review.googlesource.com/192544
Original-Reviewed-by: Tom Warren <twarren@nvidia.com>
Original-Reviewed-by: Hung-Te Lin <hungte@chromium.org>
Original-Commit-Queue: Gabe Black <gabeblack@chromium.org>
Original-Tested-by: Gabe Black <gabeblack@chromium.org>
(cherry picked from commit a672d18c3570e6991a1c1c0089697112a4cd71d0)
Signed-off-by: Marc Jones <marc.jones@se-eng.com>
Change-Id: I95e8bd7a447876174305f755cc632365ed6f5a30
Reviewed-on: http://review.coreboot.org/7734
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
Tested-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
It appears the decimal value was used instead of the hex value.
Change-Id: I04acde9e5b2a9e08ed01b0564c3d561b0385a392
Signed-off-by: Dave Frodin <dave.frodin@se-eng.com>
Reviewed-on: http://review.coreboot.org/7799
Tested-by: build bot (Jenkins)
Reviewed-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
For some of the boards using AMD processors, the Agesa Makefile.inc
is processed twice, causing the list of obj files passed to the ar
command to be added twice. This does not break the build, but does
make the ar command line unnecessarily long.
Change-Id: I02a7e6fc617e337ca2e2dceeff3d4db9995bfe16
Signed-off-by: Scott Duplichan <scott@notabs.org>
Reviewed-on: http://review.coreboot.org/7787
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
powernow_acpi.c array TDP has 20 entries, yet the loop that reads it
processes 21 entries. This causes a gcc 4.9.2 build failure. Limit
processing to 20 entries.
Change-Id: Ice173b276293184386cd8943a3213f3154f86458
Signed-off-by: Scott Duplichan <scott@notabs.org>
Reviewed-on: http://review.coreboot.org/7791
Tested-by: build bot (Jenkins)
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Remove '__attribute__((optimize("Os")))' as it is unlikely to be
necessary as it is not used in other families that have the same
code and only hides deeper issues.
Change-Id: Ica890812ebc2fb659b9c3e46b40cf3f6534b3cf2
Signed-off-by: Bruce Griffith <Bruce.Griffith@se-eng.com>
Reviewed-on: http://review.coreboot.org/7689
Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Tested-by: build bot (Jenkins)
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
The updated KaveriPI binary, upgrading to v1.1.0.7, requires changes
to define the PSP device (PCI 0:08.0) and the IOMMU device (PCI 0:00.2).
In the new AGESA binary, the IOMMU device is enabled and must be
disabled in devicetree.cb and agesawrapper_amdinitenv() to maintain
the same level of functionality.
Change-Id: I3f47e0bd5a75729ec1e4b7b11885d0622c474342
Signed-off-by: Bruce Griffith <Bruce.Griffith@se-eng.com>
Reviewed-on: http://review.coreboot.org/7727
Tested-by: build bot (Jenkins)
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Clang complains:
"implicit truncation from 'int' to bitfield changes value from -1 to 15"
-1 is define in 'c11std 6.3.1.3p2' as:
[Signed and unsigned integers] Otherwise, if the new type is unsigned,
the value is converted by repeatedly adding or subtracting one
more than the maximum value that can be represented in the new type
until the value is in the range of the new type.60)
FOOTNOTE.60 The rules describe arithmetic on the mathem...
This is "0xFF" on Mullins and "0xF" in this case. Clang seems to
complain about this two's complement in a bitfield as being truncated.
As the bitfield is 4 bits wide, (a maximum of 15 decimal), we set the
field as '0x0F'. Ideally this field /should/ be set to 'UINT8_MAX' however
we still have silly truncation warnings.
Change-Id: Ib7476d453ffd932bb911e638117cf9f56f71f269
Signed-off-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-on: http://review.coreboot.org/7719
Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Tested-by: build bot (Jenkins)
Clang complains that a signed shift result (0x210000000)
requires 35 bits to represent, but 'int' only has 32 bits.
However, we write the high bits separately and so this is
a spurious warning.
Change-Id: I3e1c57334077feb50004d7b39abff4bd84ca095b
Signed-off-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-on: http://review.coreboot.org/7673
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Tested-by: build bot (Jenkins)
They were only used internal to the SPI drivers and, according to the comment
next to their prototypes, were for when the SPI controller doesn't control the
chip select line directly and needs some help.
BUG=None
TEST=Built for link, falco, and rambi. Built and booted on peach_pit and nyan.
BRANCH=None
Original-Change-Id: If4622819a4437490797d305786e2436e2e70c42b
Original-Signed-off-by: Gabe Black <gabeblack@google.com>
Original-Reviewed-on: https://chromium-review.googlesource.com/192048
Original-Reviewed-by: Gabe Black <gabeblack@chromium.org>
Original-Tested-by: Gabe Black <gabeblack@chromium.org>
Original-Commit-Queue: Gabe Black <gabeblack@chromium.org>
(cherry picked from commit 1e2deecd9d8c6fd690c54f24e902cc7d2bab0521)
Signed-off-by: Marc Jones <marc.jones@se-eng.com>
Change-Id: Ida08cbc2be5ad09b929ca16e483c36c49ac12627
Reviewed-on: http://review.coreboot.org/7708
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
Tested-by: build bot (Jenkins)
spi_set_speed was never implemented, and spi_cs_is_valid was only implemented
as a stub and never called.
BUG=None
TEST=Built for rambi, falco, and peach_pit.
BRANCH=None
Original-Change-Id: If30c2339f5e0360a5099eb540fab73fb23582905
Original-Signed-off-by: Gabe Black <gabeblack@google.com>
Original-Reviewed-on: https://chromium-review.googlesource.com/192045
Original-Reviewed-by: Hung-Te Lin <hungte@chromium.org>
Original-Tested-by: Gabe Black <gabeblack@chromium.org>
Original-Reviewed-by: David Hendricks <dhendrix@chromium.org>
Original-Commit-Queue: Gabe Black <gabeblack@chromium.org>
(cherry picked from commit 98c1f6014c512e75e989df36b48622a7b56d0582)
Signed-off-by: Marc Jones <marc.jones@se-eng.com>
Change-Id: Iebdb2704ee81aee432c83ab182246d31ef52a6b6
Reviewed-on: http://review.coreboot.org/7707
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
Tested-by: build bot (Jenkins)
The SPI drivers for tegra and exynos5420 have code in them which waits for a
frame header and leaves filler data out. The SPI driver shouldn't have support
for frame headers directly. If a device uses them, it should support them
itself. That makes the SPI drivers simpler and easier to write.
When moving the frame handling logic into the EC support code, EC communication
continued to work on tegra but no longer worked on exynos5420. That suggested
the SPI driver on the 5420 wasn't working correctly, so I replaced that with
the implementation in depthcharge. Unfortunately that implementation doesn't
support waiting for a frame header for the EC, so these changes are combined
into one.
BUG=None
TEST=Built and booted on pit. Built and booted on nyan. In both cases,
verified that there were no error messages from the SPI drivers or the EC
code.
BRANCH=None
Original-Change-Id: I62a68820c632f154acece94f54276ddcd1442c09
Original-Signed-off-by: Gabe Black <gabeblack@google.com>
Original-Reviewed-on: https://chromium-review.googlesource.com/191192
Original-Reviewed-by: Hung-Te Lin <hungte@chromium.org>
Original-Commit-Queue: Gabe Black <gabeblack@chromium.org>
Original-Tested-by: Gabe Black <gabeblack@chromium.org>
(cherry picked from commit 4fcfed280ad70f14a013d5353aa0bee0af540630)
Signed-off-by: Marc Jones <marc.jones@se-eng.com>
Change-Id: Id8824523abc7afcbc214845901628833e135d142
Reviewed-on: http://review.coreboot.org/7706
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <gaumless@gmail.com>
Depending on the platform the underlying regions vboot requires
may not be accessible through a memory-mapped interface. Allow
for non-memory-mapped regions by providing a region request
abstraction. There is then only a few touch points in the code to
provide compile-time decision making no how to obtain a region.
For the vblocks a temporary area is allocated from cbmem. They
are then read from the SPI into the temporarily buffer.
BUG=chrome-os-partner:27094
BRANCH=None
TEST=Built and booted a rambi with vboot verification.
Original-Change-Id: I828a7c36387a8eb573c5a0dd020fe9abad03d902
Original-Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/190924
Original-Reviewed-by: Hung-Te Lin <hungte@chromium.org>
(cherry picked from commit aee0280bbfe110eae88aa297b433c1038c6fe8a3)
Signed-off-by: Marc Jones <marc.jones@se-eng.com>
Change-Id: Ia020d1eebad753da950342656cd11b84e9a85376
Reviewed-on: http://review.coreboot.org/7709
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
The arm architecture currently exports cache_sync_instructions()
in <arch/cache.h>. In order for rmodule loading to work on arm
architectures the cache_sync_instructions() needs to be called to
sequence the instruction cache. To avoid sprinkling #ifdefs around
just add an empty cache_sync_instructions() definition.
BUG=chrome-os-partner:27094
BRANCH=None
TEST=Built and booted nyan and rambi.
Original-Change-Id: I1a969757fffe0ca92754a0d953ba3630810556e3
Original-Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/191551
Original-Reviewed-by: Gabe Black <gabeblack@chromium.org>
(cherry picked from commit fda20947b928ee761d5ed15e414636af419970a6)
Signed-off-by: Marc Jones <marc.jones@se-eng.com>
Change-Id: I3e8ca12e1d82ccedf1ff9851ae3c5c80cda2dd5f
Reviewed-on: http://review.coreboot.org/7710
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
To find the coreboot tables, the payload has historically searched for their
signature in a predefined region of memory. This is a little clumsy on x86,
but it works because you can assume certain regions are RAM. Also, there are
areas which are set aside for the firmware by convention. On x86 there's a
forwarding entry which goes in one of those fairly small conventional areas
and which points to the CBMEM area at the end of memory.
On ARM there aren't areas like that, so we've left out the forwarding entry and
gone directly to CBMEM. RAM may not start at the beginning of the address space
or go to its end, and that means there isn't really anywhere fixed you can put
the coreboot tables. That's meant that libpayload has to be configured on a
per board basis to know where to look for CBMEM.
Now that we have boards that don't have fixed amounts of memory, the location
of the end of RAM isn't fixed even on a per board level which means even that
workaround will no longer cut it.
This change makes coreboot pass the location of the coreboot tables to
libpayload using r0, the first argument register. That means we'll be able to
find them no matter where CBMEM is, and we can get rid of the per board search
ranges.
We can extend this mechanism to x86 as well, but there may be more
complications and it's less necessary there. It would be a good thing to do
eventually though.
BUG=None
TEST=Built and booted on nyan. Changed the size of memory and saw that the
payload could still find the coreboot tables where before it couldn't. Built
for pit, snow, and big.
BRANCH=None
Original-Change-Id: I7218afd999da1662b0db8172fd8125670ceac471
Original-Signed-off-by: Gabe Black <gabeblack@google.com>
Original-Reviewed-on: https://chromium-review.googlesource.com/185572
Original-Reviewed-by: Julius Werner <jwerner@chromium.org>
Original-Commit-Queue: Gabe Black <gabeblack@chromium.org>
Original-Tested-by: Gabe Black <gabeblack@chromium.org>
(cherry picked from commit ca88f39c21158b59abe3001f986207a292359cf5)
Signed-off-by: Marc Jones <marc.jones@se-eng.com>
Change-Id: Iab14e9502b6ce7a55f0a72e190fa582f89f11a1e
Reviewed-on: http://review.coreboot.org/7655
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
The preceding patch copied gizmo2 from the amd/olivehill
board. This commit includes the changes required to make
the code reflect the gizmo2 hardware:
- Update the vendor Kconfig to add gizmo2
- Update the mainboard Kconfig
- Update devicetree
- Add support in for the soldered down DDR3
- Update the CODEC verb data
- Update the graphics connector settings
- Adjust the temperature thresholds for the fan
What's missing:
- Interrupt routing tables
Gizmo2 can boot DOS and Ubuntu 14.10.
Change-Id: I3d7202957c082974689f2a8c04d8cd33dbdc1a89
Signed-off-by: Dave Frodin <dave.frodin@se-eng.com>
Reviewed-on: http://review.coreboot.org/7722
Tested-by: build bot (Jenkins)
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Clang warns of an implicit conversion from 'double' to 'int'
e.g. changes value from '26.67' to '26'. Thus take the floor() of
the array and not change orginal behaviour.
Change-Id: Ifcc7bbfe8d627451b82053f53a885f315e2550ec
Signed-off-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-on: http://review.coreboot.org/7725
Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Tested-by: build bot (Jenkins)
Correct mask to select bits 4-6 inclusively as per comment and use
bitwise operations while working with bits. Be sure to write back out
the data on the retrain.
See:
commit cab9efb2 southbridge/amd/rsXY0/cmn.c: Fix bitwise logic and mask in loop
Change-Id: I95d1799514157b7849f3e473837aaf2fd9bd59b9
Signed-off-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-on: http://review.coreboot.org/7692
Reviewed-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Tested-by: build bot (Jenkins)
- In '-ffreestanding' main() is just as any other function and so
it needs a type-signature. Fixes a clang warning.
- Bay Trail and Rangeley have the updated romstage.c with the code
moved into the chipset, put the prototype in romstage.c.
- The sandybridge code has not been updated, so the prototype
for it goes into chipset_fsp_util.h, next to the prototype for
romstage_main_continue.
- Correct the return value of baytrail main() from void * to void
and remove the unnecessary asmlinkage tag. I'm surprised that this
didn't generate a warning...
Change-Id: I85ac0797d1e55d2b7ffdca039a52820d7827e704
Signed-off-by: Martin Roth <martin.roth@se-eng.com>
Reviewed-on: http://review.coreboot.org/7724
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marc.jones@se-eng.com>
This is a direct copy of the amd/olivehill mainboard which
will be the starting point for this port.
Change-Id: I6a643f7ac35d89e21df0ffdf4e61a2da46e19b82
Signed-off-by: Dave Frodin <dave.frodin@se-eng.com>
Reviewed-on: http://review.coreboot.org/7721
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Tested-by: build bot (Jenkins)
This is not actually required. Tested on 'minnow max' hardware as
well as compared the asm of the optimized and non-optimized. Thanks Martin!
Change-Id: I06e71876c3a3a15101013623797c2ebbf449756d
Signed-off-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Found-by: Clang
Reviewed-on: http://review.coreboot.org/7694
Reviewed-by: Martin Roth <gaumless@gmail.com>
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
If it's a 4 byte format (as per documentation), there
are some reserved bits, so let's mark them as such...
Change-Id: I50f12cfff2c9bb9d082a5f3c3ac54c0d514d862b
Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Reviewed-on: http://review.coreboot.org/7674
Tested-by: build bot (Jenkins)
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Following the same reasoning as commit
ee905a8 vendorcode/amd/agesa/fam15tn: Build as a static library
Since AGESA is stage-independent, we can build it just once, and use
the resulting static library in both rom and ram stages.
Change-Id: I8fbb318daacf64a14a71022705eb040a01c34fa8
Signed-off-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-on: http://review.coreboot.org/7699
Tested-by: build bot (Jenkins)
Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Following the same reasoning as commit
ee905a8 vendorcode/amd/agesa/fam15tn: Build as a static library
Since AGESA is stage-independent, we can build it just once, and use
the resulting static library in both rom and ram stages.
Change-Id: I7798b689db3e582649eb4af4ccd1877bb1d49063
Signed-off-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-on: http://review.coreboot.org/7698
Tested-by: build bot (Jenkins)
Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Following the reasoning in,
8089f17 mainboard/lenovo/x230 Fix usage of GNU field designator extension
In C99 we defined a syntax for this. GCC's old syntax was deprecated.
Change-Id: Id3b16872f62660393d938d6f95977a4e3842d0d1
Signed-off-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-on: http://review.coreboot.org/7690
Tested-by: build bot (Jenkins)
Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Silence some useless Clang warns in this case.
Change-Id: I202a85f7dec52c65d80e2bc56f7d9e4eb3e61d48
Signed-off-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-on: http://review.coreboot.org/7696
Tested-by: build bot (Jenkins)
Reviewed-by: Vladimir Serbinenko <phcoder@gmail.com>
Its scope is limited to a single mainboard and is only to go through ifdef.
Kill it and move the value to the code.
Change-Id: I76a87e2790d57dee8f37b51e33d0689fffd3a59d
Signed-off-by: Vladimir Serbinenko <phcoder@gmail.com>
Reviewed-on: http://review.coreboot.org/7135
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Currently due to enum mistake DDR3 = 0xb was confused with DIMM type and
interpreted as LRDIMM, considered unregistered and so every RAM was
unregistered.
Registered RAM is rarely used, so I suppose the code was never tested with them.
For unregistered RAM exactly the same codepath is followed.
Change-Id: I02fe8b1fd7be3bd382399ffa0eb513965a2a6d77
Signed-off-by: Vladimir Serbinenko <phcoder@gmail.com>
Reviewed-on: http://review.coreboot.org/7687
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Tested-by: build bot (Jenkins)
I have no such board to check the real fixes but this board shouldn't block
benefits for the rest of the tree.
Change-Id: I9e9d4af1b360bcf0099ac2901b08f7fcd7569097
Signed-off-by: Vladimir Serbinenko <phcoder@gmail.com>
Reviewed-on: http://review.coreboot.org/7681
Tested-by: build bot (Jenkins)
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Keep only the last one: it was the one which was really used.
Change-Id: I19132f6224d6847e615e3c582aaa6e66b0d56c7a
Signed-off-by: Vladimir Serbinenko <phcoder@gmail.com>
Reviewed-on: http://review.coreboot.org/7677
Tested-by: build bot (Jenkins)
Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Topology Extensions Support (bit 54 of 0xC0011005) applies to
PACKAGE_TYPE_FS1r2 also. Rids us of:
"Re-enabling disabled Topology Extensions Support"
showing up in dmesg.
Change-Id: Id123fa9632936c150cf1aebc4d34b404a4398ead
Signed-off-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-on: http://review.coreboot.org/7671
Tested-by: build bot (Jenkins)
Reviewed-by: Bruce Griffith <Bruce.Griffith@se-eng.com>
The current interrupt routing shares interrupt 5 between LPC and PCI which
isn't possible.
Use IRQ 11 for all devices in PCI mode. Move conflicting LPC to free IRQ.
Change-Id: I3ac8c2f19195ef6b07f4ee7dde64dd038d024126
Signed-off-by: Vladimir Serbinenko <phcoder@gmail.com>
Reviewed-on: http://review.coreboot.org/7477
Tested-by: build bot (Jenkins)
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
This config is used only to generate PIRQ table. If no such table is
supplied there is no need for config.
Change-Id: I537d440f53019a6bf7f190446074e75e7420545a
Signed-off-by: Vladimir Serbinenko <phcoder@gmail.com>
Reviewed-on: http://review.coreboot.org/7566
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Tested-by: build bot (Jenkins)
The contents of these files were guarded by a check for the _MSC_VER
macro, which we don't use.
Change-Id: Ic595c8e6284c54e1449cf21e0cebee8c9ce7c682
Signed-off-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Reviewed-on: http://review.coreboot.org/7670
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Tested-by: build bot (Jenkins)
Don't build in non-essential blobs by default. However, if the user
selected to use the blobs repository, then default to including the
blobs.
Change-Id: Ie90f00d7c18d725f24fe1503fadaf098d3cefa4a
Signed-off-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Signed-off-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Reviewed-on: http://review.coreboot.org/7638
Tested-by: build bot (Jenkins)
Rather than have Linux report:
i8042: PNP: No PS/2 controller found. Probing directly.
and go off probing PNP config space, build in EC ASL for the
PS/2 keyboard and mouse.
The ASL explicitly passes these resources to the Linux to avoid
said probe.
ASL Details:
PS/2 keyboard (PNP0303 at 0x60,0x64 irq 1 )
PS/2 mouse (PNP0F13 at 0x60,0x64 irq 12)
Change-Id: I0697fab65915907fbe2b3551182b3a1b0d665ddb
Signed-off-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-on: http://review.coreboot.org/7651
Tested-by: build bot (Jenkins)
Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Provide ASL to support the AUX port (a.k.a Mouse) found at
0x60,0x64 irq 12 on this EC.
Change-Id: I6969ae4d492570136a8e14e42509638857e1ed85
Signed-off-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-on: http://review.coreboot.org/7650
Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Tested-by: build bot (Jenkins)
Toggle on in devicetree.cb and build into AGESA by buildOpts.c.
Add ACPI and MPTABLES interrupt routers for IOMMU also.
Change-Id: Ia838f9b70f09ed1180daeb5382edc08c4b74946c
Signed-off-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-on: http://review.coreboot.org/7643
Tested-by: build bot (Jenkins)
Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Missing IOMMU support is missing from the libagesa Makefile, it also
lacks a header with type-signature and a few bad typecast issues.
Change-Id: I7f2ad2104de9baaa66dbb6ffeb0f2b4d35fa5c16
Signed-off-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Co-Author: Rudolf Marek <r.marek@assembler.cz>
Reviewed-on: http://review.coreboot.org/7642
Tested-by: build bot (Jenkins)
Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Use non-local inclusion syntax over relative paths for
'drivers/pc80/ps2_controller.asl'.
Change-Id: Ie2bfa893dc268ec5118d2a9addadbc759d85d357
Signed-off-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-on: http://review.coreboot.org/7664
Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Tested-by: build bot (Jenkins)
Right now, coreboot code using AGESA headers can only build if all the
AGESA path are given to the compiler via the "-I" option. This is sub-
optimal, as it requires us to have every AGESA source directory
specified as a compiler include path. This pollutes our global include
paths.
We restrict the compiler include paths to only allow "AGESA_ROOT/" and
"AGESA_ROOT/Include". We then modify the AGESA headers to specify
non-local include files relative to "AGESA_ROOT/Include".
We use the convention that includes relative to the directory of the
header are included as "path/to/header.h", while includes relative to
AGESA_ROOT are included as <path/to/header.h>.
This change allows building coreboot code based on AGESA with the
limited subset of include paths, but does not allow AGESA itself to
build with this restricted subset.
Change-Id: I31102273c8caa8d6b1d80774bfd35711825bec03
Signed-off-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Reviewed-on: http://review.coreboot.org/5424
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Tested-by: build bot (Jenkins)
Minor fix to avoid confusion, nothing to see here.
Change-Id: I89d56a91d2df049e85cf49c23218620caba84880
Signed-off-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-on: http://review.coreboot.org/7654
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
It was requested to be able to update XHCI vs EHCI via get_option,
so I've added it here for minnow max. This could get moved to the
chipset_fsp_util.c file later, but I'm adding it here for now.
More checking needs to be added to this:
- Are both controllers enabled in devicetree? If not, we don't want
to allow the switch.
Change-Id: I4d8d2229cb9fa0cd9068701454b28ffac6d8e767
Signed-off-by: Martin Roth <martin.roth@se-eng.com>
Reviewed-on: http://review.coreboot.org/7633
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Marc Jones <marc.jones@se-eng.com>
- Align register values.
- Enable both EHCI and XHCI so the choice of port used can be made
at runtime. When both are enabled in devicetree, XHCI currently gets
disabled by the FSP chipset code. This can be overridden in mainboard
code or by a Kconfig entry, but there's a question about whether or not
that's desired.
- Enable function 1c.0 so the rest of the functions will be
seen, even though the function is not actually used. This is a
short-term fix, as the correct solution is to determine whether or not
any of the other functions are enabled, and not to hide function 0 if
they are. I am working on that, but I want to get this in for now.
Change-Id: I83ae12c2393024b82a55d0b3a5ffa8782e16107e
Signed-off-by: Martin Roth <martin.roth@se-eng.com>
Reviewed-on: http://review.coreboot.org/7663
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marc.jones@se-eng.com>
- The EDS has the function disable bit for eMMC incorrectly listed
as 8. Changing it back to the correct bit 11.
- The FSP will disable functions that it is told are disabled, so
coreboot code that disables the functions is redundant. Removing it.
Change-Id: I95c31d92d3af5182ddf7fd47f651bbb61cdedb82
Signed-off-by: Martin Roth <martin.roth@se-eng.com>
Reviewed-on: http://review.coreboot.org/7653
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marc.jones@se-eng.com>
- The ROM chip is 8MB, not 4MB.
- Default to the 2GB SKU instead of 1GB - that's what's out right now.
- Set CBFS size to 3MB - that's what the firmware descriptor is set to.
Change-Id: Ic77f5c1e898dca39de573623707ff5f5e5ca9682
Signed-off-by: Martin Roth <martin.roth@se-eng.com>
Reviewed-on: http://review.coreboot.org/7649
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marc.jones@se-eng.com>
The documentation for the FSP gives the name as BAYTRAIL_FSP.fd instead
of the old FvFsp.bin.
Change-Id: I69c7c5ff49afd6552612cf50c9ca9b30cfb003e2
Signed-off-by: Martin Roth <martin.roth@se-eng.com>
Reviewed-on: http://review.coreboot.org/7648
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marc.jones@se-eng.com>
New microcode for Bay Trail I B2/B3 and D0 parts was released in the
Gold 3 Bay Trail FSP release.
Change the microcode size to an area instead of the exact size of the
patches. This will hopefully reduce updates to the microcode size.
Change-Id: I58b4c57a4bb0e478ffd28bd74a5de6bb61540dfe
Signed-off-by: Martin Roth <martin.roth@se-eng.com>
Reviewed-on: http://review.coreboot.org/7647
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marc.jones@se-eng.com>
Move the Kconfig variable into a .h file - this does not need to be
in Kconfig.
Change-Id: I1db20790ddb32e0eb082503c6c60cbbefa818bb9
Signed-off-by: Martin Roth <martin.roth@se-eng.com>
Reviewed-on: http://review.coreboot.org/7646
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marc.jones@se-eng.com>
Include clock.c in the appropriate coreboot stages, modify the code to
build cleanly. Use proper pointer cast in .h files.
BUG=chrome-os-partner:27784
TEST='emerge-storm coreboot' still succeeds
Original-Change-Id: I227c871b17e571f6a1db3ada3821dbb1ee884e59
Original-Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/196407
(cherry picked from commit 75decceccd97298974891bb98b796eccfe11f46c)
Signed-off-by: Marc Jones <marc.jones@se-eng.com>
Change-Id: I7d44464d4ca8153e84407fc05a25e2e79e74901e
Reviewed-on: http://review.coreboot.org/7271
Tested-by: build bot (Jenkins)
Reviewed-by: David Hendricks <dhendrix@chromium.org>
These driver needs to be in src/lib, and the include file needs to be
renamed to avoid collision with the top level uart.h.
BUG=chrome-os-partner:27784
TEST=emerge-storm coreboot still works
Original-Change-Id: Ie12f44e055bbef0eb8b1a3ffc8d6742e7a446942
Original-Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/196393
(cherry picked from commit c5618fd418642f5b009582f5f6bc51f7c9d54bec)
Signed-off-by: Marc Jones <marc.jones@se-eng.com>
Change-Id: I5e25ae350ac5e71b47a0daef078b03cc5ac35401
Reviewed-on: http://review.coreboot.org/7270
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
Set the UPD entry based on the Kconfig value instead of having two
separate places that the value needs to be set.
Change-Id: I3d32111b59152d0a8fc49e15320c7b5a140228a6
Signed-off-by: Martin Roth <martin.roth@se-eng.com>
Reviewed-on: http://review.coreboot.org/7490
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marc.jones@se-eng.com>
Reviewed-by: FEI WANG <wangfei.jimei@gmail.com>
Update the printk statements to use FSP_INFO_LEVEL instead of
BIOS_DEBUG. These values are currently identical, but by using the
second #define, it lets them all be changed as a unit. This can
be overridden for a particular platform by adding a #define in
chipset_fsp_util.c.
Change-Id: Idbf7e55090230ec940c7c8cd3ec8632461561428
Signed-off-by: Martin Roth <martin.roth@se-eng.com>
Reviewed-on: http://review.coreboot.org/7520
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marc.jones@se-eng.com>
- Update chipset_fsp_util.c to use the UPD_DEVICE_CHECK macro. This
makes the code more standardized and easier to read.
- Add some debug printing that was removed in the transition.
Change-Id: Iea24dd9ca53f39791bc6371291a3fa7a6fc5ed0f
Signed-off-by: Martin Roth <martin.roth@se-eng.com>
Reviewed-on: http://review.coreboot.org/7498
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marc.jones@se-eng.com>
- Update chipset_fsp_util.h to add the UPD_MEMDOWN_CHECK pointing into
the PcdMemoryParameters structure. This is baytrail FSP specific, so
it's put into the chipset code instead of the 'driver' code. Since some
of the values need to be decremented and some do not, a second parameter
was added to control this. This macro also does not print out the
values as they are printed out separately if memory down is enabled.
- Update chipset_fsp_util.c to use the UPD_MEMDOWN_CHECK macro. This
makes the code more standardized and easier to read.
Change-Id: I233e45db43af4726cab41f4880f1706cf8abb0b7
Signed-off-by: Martin Roth <martin.roth@se-eng.com>
Reviewed-on: http://review.coreboot.org/7632
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marc.jones@se-eng.com>
Update chipset_fsp_util.c to use the UPD_SPD_CHECK macro. This
makes the code more standardized and easier to read.
Change-Id: I9944e1a4df82e64a205598e98ed0f3b840af1019
Signed-off-by: Martin Roth <martin.roth@se-eng.com>
Reviewed-on: http://review.coreboot.org/7489
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
Tested-by: build bot (Jenkins)
- Update chipset_fsp_util.c to use the UPD_DEFAULT_CHECK macro. This
makes the code more standardized and easier to read.
- Update chip.h to use standardized macros
Change-Id: Icbe5ec92b0aa31e21f3dd1593a96b246d83008f7
Signed-off-by: Martin Roth <martin.roth@se-eng.com>
Reviewed-on: http://review.coreboot.org/7488
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marc.jones@se-eng.com>
Add macros and #defines for working with the UPD data. This makes
the code look much cleaner.
Remove the UPD_ENABLE / UPD_DISABLE from fsp_rangeley/chip.h and include
the fsp_values header instead. This fixes a conflict.
Change-Id: I72c9556065e5c7461432a4593b75da2c8a220a12
Signed-off-by: Martin Roth <martin.roth@se-eng.com>
Reviewed-on: http://review.coreboot.org/7487
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marc.jones@se-eng.com>
Add a section .illegal_globals to romstage and check that the section does not
contain any variables while creating romstage.
[pg: Handle individual AGESA special cases in the
linker script instead of whitelisting everything
remotely AGESA related in the Makefile.]
Change-Id: I866681f51a44bc21770d32995c281b556a90c153
Signed-off-by: Furquan Shaikh <furquan@google.com>
Signed-off-by: Patrick Georgi <pgeorgi@google.com>
Reviewed-on: http://review.coreboot.org/7306
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
TL;DR ASCII art that sucks, remove it.
Change-Id: I424736b040fe019bba6155de76903225a266760d
Signed-off-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-on: http://review.coreboot.org/7641
Tested-by: build bot (Jenkins)
Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Since this board does not provide a PIRQ table.
Change-Id: I1068dd99c4cecdd2113484fe24ae2bb86a058cb3
Signed-off-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-on: http://review.coreboot.org/7644
Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Tested-by: build bot (Jenkins)
This makes lzmadecode 64-bit clean (I hope).
It also cleans up a few other nits.
Change-Id: I24492e9f357e8d3a6de6abc351267f900eb4a19a
Signed-off-by: Ronald G. Minnich <rminnich@gmail.com>
Reviewed-on: http://review.coreboot.org/7623
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
Tested-by: build bot (Jenkins)
We don't actually use nor support these as our implementation
makes use of gcccar.inc. They maybe useful as a reference for
history so lets keep them in version history.
Change-Id: I388251dead449dde14283e57db39c37982d947b2
Signed-off-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-on: http://review.coreboot.org/7596
Tested-by: build bot (Jenkins)
Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Otherwise checksum may not work correctly on early stages.
For compatibility with old bootblocks also enable it early in romstage.
Change-Id: Ie541d71bd76af182e445aa5ef21fe5ba77091159
Signed-off-by: Vladimir Serbinenko <phcoder@gmail.com>
Reviewed-on: http://review.coreboot.org/7556
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
The Embedded Controller sits behind the LPC bridge and so needs
LPC decodes to be enabled.
Remove the LPC decode enable out of agesawrapper.c. The enable
is in fact done in: 'VOID FchInitResetLpcProgram(IN VOID *FchDataPtr)'
which writes the magic '0xFF03FFD5' to register 0x44 of the PCI 14.3
LPC Bridge to enable LPC decodes when HUDSON_LEGACY_FREE is not defined.
Change-Id: Ia487d21faa0fceb2557dbce14ef8822116fada91
Signed-off-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-on: http://review.coreboot.org/7628
Reviewed-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Tested-by: build bot (Jenkins)
The Embedded Controller sits behind the LPC bridge and so needs
LPC decodes to be enabled.
Remove the LPC decode enable out of agesawrapper.c. The enable
is in fact done in: 'VOID FchInitResetLpcProgram(IN VOID *FchDataPtr)'
which writes the magic '0xFF03FFD5' to register 0x44 of the PCI 14.3
LPC Bridge to enable LPC decodes when HUDSON_LEGACY_FREE is not defined.
Change-Id: I0b4e99cc0d6f89f0261f26ee61b8c175a373c730
Signed-off-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-on: http://review.coreboot.org/7625
Reviewed-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Tested-by: build bot (Jenkins)
Enabling MMCONF PCI-e configuration access should be done before
console_init(). This will likely move further to bootblock one day.
Change-Id: I20c93fe6e79ef7e7981b2f1cd3c6b446feea0f4e
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: http://review.coreboot.org/7163
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Tested-by: build bot (Jenkins)
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Gabe Black