The amdht code currently relies on an idiosyncratic ASSERT() macro,
which actually doesn't do anything right now, and even it did would only
print a janky error message. Replace this with the normal ASSERT() macro
from <assert.h>. The default behaviour now is to print an error message
but do nothing else, and failed assertions will only halt if you enable
FATAL_ASSERT, in which case, well, you asked for it.
Change-Id: I6db7565171a345f9afbc9fb37cff8fda58f942df
Signed-off-by: Jacob Garber <jgarber1@ualberta.ca>
Found-by: Coverity CID 1402076
Reviewed-on: https://review.coreboot.org/c/coreboot/+/34375
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
- Use log2() when rounding down size_mb to the closest power of 2.
Do a sanity check beforehand that size_mb is nonzero, else log2()
will return -1 and there will be an undefined integer shift.
- The framebuffer size needs to be between 8 and 512 MiB, so check
after all the calculations are done to make sure this is the case.
Change-Id: I3962e5cdc094c8da22d8dbadf16637e02fa98689
Signed-off-by: Jacob Garber <jgarber1@ualberta.ca>
Found-by: Coverity CID 1391086
Reviewed-on: https://review.coreboot.org/c/coreboot/+/34355
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
This switch was likely copy-pasted from the one right above it. However,
the MEM_CLOCK_800MHz case isn't needed, since that is explicitly checked
and avoided before the while loop. With that gone, only the
667MHz/default case is left, which we don't need to switch over anymore.
Change-Id: Idfb9cc27dd8718f627d15ba92a9c74c51c2c1c2d
Signed-off-by: Jacob Garber <jgarber1@ualberta.ca>
Found-by: Coverity CID 1347372
Reviewed-on: https://review.coreboot.org/c/coreboot/+/33407
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: HAOUAS Elyes <ehaouas@noos.fr>
Reviewed-by: David Hendricks <david.hendricks@gmail.com>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Read the northbridge BARs from device PCI0:0.0.
Untested.
Change-Id: I27bfb5721d9ae3dc5629942ebac29b12a7308441
Signed-off-by: Patrick Rudolph <patrick.rudolph@9elements.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/32074
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
Here, vga_disable stored but we read it only if MAINBOARD_DO_NATIVE_VGA_INIT.
Found-by: scan-build 7.0.1-8
Change-Id: I5c359df71568b56f48eca9615c6265da33d4a073
Signed-off-by: Elyes HAOUAS <ehaouas@noos.fr>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/34331
Reviewed-by: Patrick Rudolph <siro@das-labor.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
The GCFC (Graphics Clock Frequency Control) read is not used at
the line below.
As the default value is zero, let's remove unused read.
Found-by: scan-build 7.0.1-8
Change-Id: I82c567e3a5b0c0c4a8596ea0cb7693667c71b720
Signed-off-by: Elyes HAOUAS <ehaouas@noos.fr>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/34329
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
The speed argument should be one of the six values from the mem_clock
enum, so something is very wrong if this is not the case. Better to
die now than return 0, which will cause a division-by-zero error
later on where this function is called. The first two speeds are also
unsupported and have the same problem with returning 0, so die on those
as well.
Change-Id: Ib628c0eed3d6571bdde1df27ae213ca0691ec256
Signed-off-by: Jacob Garber <jgarber1@ualberta.ca>
Found-by: Coverity CID 1391088
Reviewed-on: https://review.coreboot.org/c/coreboot/+/33409
Reviewed-by: David Hendricks <david.hendricks@gmail.com>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-by: HAOUAS Elyes <ehaouas@noos.fr>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
If the decoded SPD DRAM frequency is slower than the controller minimum,
then there will be an unsigned integer underflow in the following loop,
which will lead to a very large out of bounds array access. Ensure this
does not happen.
Change-Id: Ic8ed1293adfe0866781bd638323977abd110777e
Signed-off-by: Jacob Garber <jgarber1@ualberta.ca>
Found-by: Coverity CID 1229675
Reviewed-on: https://review.coreboot.org/c/coreboot/+/33383
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: HAOUAS Elyes <ehaouas@noos.fr>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
In our codebase, this is only coupled with intel/e7505.
The PCI registers reference here were for intel/i945.
Also aseg_smm_lock() was previously not called.
Change-Id: I21d991c8c2f5c2dde1f148fd80963e39d9836d3c
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/34149
Reviewed-by: Arthur Heymans <arthur@aheymans.xyz>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
- Remove unnecessary braces
- Move variable assignment out of function call
- Do not find lowest bit set of 0, which is undefined
- Use unsigned integer when bit shifting
Change-Id: I8651f8cd04165d8d31c44f7919ad5e43499d3d4c
Signed-off-by: Jacob Garber <jgarber1@ualberta.ca>
Found-by: Coverity CID 1229562
Reviewed-on: https://review.coreboot.org/c/coreboot/+/33381
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: HAOUAS Elyes <ehaouas@noos.fr>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Die if there are no memory ranks found to prevent a division by zero.
Change-Id: I6146dd8420f3734d1a672a9f29a098f47fcb739c
Signed-off-by: Jacob Garber <jgarber1@ualberta.ca>
Found-by: Coverity CID 1229628
Reviewed-on: https://review.coreboot.org/c/coreboot/+/33403
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: HAOUAS Elyes <ehaouas@noos.fr>
Reviewed-by: Nico Huber <nico.h@gmx.de>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
This is only a qualifier between TSEG and ASEG.
Change-Id: I8051df92d9014e3574f6e7d5b6f1d6677fe77c82
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/34135
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Furquan Shaikh <furquan@google.com>
While common to many Intel CPUs, this is not an architectural
MSR that should be globally defined for all x86.
Change-Id: Ibeed022dc2ba2e90f71511f9bd2640a7cafa5292
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/34090
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: HAOUAS Elyes <ehaouas@noos.fr>
Reviewed-by: David Guckian
With VBOOT=y && VBOOT_MEASURED_BOOT=y message
digest will be allocated from the stack and
1 KiB reserve used with the recent platforms
was no longer sufficient.
The comment of LZMA scratchpad consuming stack
was obsolete for postcar, so these can be reduced
to same 4 KiB.
Change-Id: Iba1fb5bfad6946f316feac2d8c998a782142a56a
Signed-off-by: Mario Scheithauer <mario.scheithauer@siemens.com>
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/33775
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-by: Werner Zeh <werner.zeh@siemens.com>
It's forbidden to use dereference CAR_GLOBAL variables
directly. The notation fails after CAR teardown for
romstage.
Change-Id: I6e6285ca0f520608c2a344517fbac943aeb36d87
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/33995
Reviewed-by: Arthur Heymans <arthur@aheymans.xyz>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Only (conditionally) used part was dump_pci_device()
and that was never particularly useful either.
Change-Id: Iaacfa511de1ce1e0bdbd2e8a74e41d336e505670
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/33958
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Arthur Heymans <arthur@aheymans.xyz>
This function is duplicated in many AMD northbridge files, and all
the definitions have started to diverge somewhat. This moves a single
copy into device utils and deletes the rest. The function definition
from nb/amd/amdfam10 was chosen to be kept, since it contains several
fixes from commit 59d609217b (AMD fam10: Fix add_more_links) that
the others don't have.
For the ease of diffing, the checkpatch lints and other small cleanups
will be done in a follow-up patch.
Change-Id: I5ea73126092449db3f18e19ac0660439313072de
Signed-off-by: Jacob Garber <jgarber1@ualberta.ca>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/33237
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
All the clocks are switched on anyway, so this series of if statements
isn't needed.
Change-Id: I654043fd6736caa6890fd697015c577ddaa7cd41
Signed-off-by: Jacob Garber <jgarber1@ualberta.ca>
Found-by: Coverity CID 13473{27-30}
Reviewed-on: https://review.coreboot.org/c/coreboot/+/33408
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: HAOUAS Elyes <ehaouas@noos.fr>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
As described in the comment, the BKDG-recommended algorithm causes
problems and isn't being used, so just delete it.
Change-Id: Ib3d4eba7ea2d7f6545613af5670b243bc011275f
Signed-off-by: Jacob Garber <jgarber1@ualberta.ca>
Found-by: Coverity CID 1347325, 1347326
Reviewed-on: https://review.coreboot.org/c/coreboot/+/33269
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
The literal needs to be cast to a uint64 to prevent num_nodes
from being implicitly promoted to a signed int.
Change-Id: Id2fa328fb8d0a9827c7c78157c024736e9b26dc4
Signed-off-by: Jacob Garber <jgarber1@ualberta.ca>
Found-by: Coverity CID 1347343
Reviewed-on: https://review.coreboot.org/c/coreboot/+/33210
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Martin Roth <martinroth@google.com>
Reviewed-by: HAOUAS Elyes <ehaouas@noos.fr>
The case when Speed == 5 is already dealt with on line 111.
Change-Id: Icc41c00fb333a51001568f588c17f7b6c6a3a923
Signed-off-by: Jacob Garber <jgarber1@ualberta.ca>
Found-by: Coverity CID 1229626
Reviewed-on: https://review.coreboot.org/c/coreboot/+/33213
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: HAOUAS Elyes <ehaouas@noos.fr>
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
This check is already performed at the beginning of
dct_ddr_voltage_index().
Change-Id: Ia73025333c152f14249afb16d05f66791e69e7cb
Signed-off-by: Jacob Garber <jgarber1@ualberta.ca>
Found-by: Coverity CID 1347322, 1347323
Reviewed-on: https://review.coreboot.org/c/coreboot/+/33264
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
This conditional is exactly the same as the one right below it, except
the operations are reorganized slightly.
Change-Id: I00c19a467d23a0736bc2a33b516f97080039e634
Signed-off-by: Jacob Garber <jgarber1@ualberta.ca>
Found-by: Coverity CID 1347321
Reviewed-on: https://review.coreboot.org/c/coreboot/+/33267
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
The sandybridge systemagent-r6 blob is modified:
- To be more flexible about the location of the stack w.r.t. the heap
- Place the MRC pool right below the MRC_VAR region
- to work with the same DCACHE_RAM_BASE from the native raminit (could
make the CAR linker symbols easily compatible if desired)
This allows CAR setup compatibility between mrc.bin and native
bootpath and also allows for BIOS/memory mappeds region larger than
8MB.
This changes the semantics of CONFIG_DACHE_RAM_MRC_VAR_SIZE to also
include the pool on top of MRC_VAR region.
TESTED on T520 (boots and resumes from S3 with mrc.bin).
Change-Id: I17d240656575b69a24718d90e4f2d2b7339d05a7
Signed-off-by: Arthur Heymans <arthur@aheymans.xyz>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/33228
Reviewed-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-by: Nico Huber <nico.h@gmx.de>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
These if statements can be combined to merge the two branches of the
conditional and remove the duplicate pDCTstat->Speed == 3 check.
Change-Id: I41aa19b4b7ed7b1a0e4f83f72e66869760e677dd
Signed-off-by: Jacob Garber <jgarber1@ualberta.ca>
Found-by: Coverity CID 1229583
Reviewed-on: https://review.coreboot.org/c/coreboot/+/33211
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
These two functions try to access arrays of lengths 32 and 64 at indices
of at most 259 and 71 (respectively). Something here is seriously wrong.
This code was introduced in 2007, and aside from cosmetic changes, has
had no modifications since then. I don't know what this code is supposed
to do, and asking around on IRC, no one else did either. Until someone
has the interest and time to work on it, let's at least add a die() to
prevent the out of bounds access and alert the user that something is
wrong.
Change-Id: I5fc15a50a9f0e97add31e3a40da82a15f7427358
Signed-off-by: Jacob Garber <jgarber1@ualberta.ca>
Found-by: Coverity CID 12296{79-82}
Reviewed-on: https://review.coreboot.org/c/coreboot/+/33404
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Arthur Heymans <arthur@aheymans.xyz>
* Move the contents of iommu.c to early_init.c.
* Name the functions like done in intel/soc/common.
* Move PAMx register setup to own function
Preparations for integration in soc/intel/common/*
Tested on Lenovo T520 (Intel Sandy Bridge).
Still boots to OS, no errors visible in dmesg.
Change-Id: I3ec395bf6722bceb84316e92733dcfcd7a093639
Signed-off-by: Patrick Rudolph <patrick.rudolph@9elements.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/32068
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
