This change adds and updates headers in all of the northbridge files
that had missing or unrecognized headers. After this goes in, we can
turn on lint checking for headers in all northbridge directories.
Change-Id: I8cd7c04ddb8e58946dcdf9c7c125e23698647a73
Signed-off-by: Martin Roth <martinroth@google.com>
Reviewed-on: https://review.coreboot.org/26571
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Werner Zeh <werner.zeh@siemens.com>
When programming the final dram attribute and dram boundary settings,
on DDR3 dram one also needs to enable ZQCAL in the CxREFRCTRL (DRAM
Refresh Control) register as documented in "Intel ® 4 Series Chipset
Family" documentation.
Change-Id: I11a79f6800dbfe19c2bd33c0d6caca14b034e384
Signed-off-by: Arthur Heymans <arthur@aheymans.xyz>
Reviewed-on: https://review.coreboot.org/22996
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Patrick Rudolph <siro@das-labor.org>
DDR3 adapted a fly-by topology which allows for better signal
integrity but at the same time requires additional calibration. This
is done by settings the targeted rank in write leveling mode while
disabling output buffer on the other ranks. After that the DQS signal
gets sampled over DQ until a transition from high to low is found.
Change-Id: I695969868b4534f87dd1f37244fdfac891a417f0
Signed-off-by: Arthur Heymans <arthur@aheymans.xyz>
Reviewed-on: https://review.coreboot.org/22995
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
Add DDR3 JEDEC init (Power up and Initialization by setting emrs regs)
This also modifies the send_jedec_cmd function as DDR3 dimms can have
ranks mirrored which needs to be accounted for.
The ddr3_emrs1_config array is placed externally since it is also
needed for write leveling.
Change-Id: I510b8669aaa48ba99fb4dcf1ece716aef26741bb
Signed-off-by: Arthur Heymans <arthur@aheymans.xyz>
Reviewed-on: https://review.coreboot.org/22994
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
Stack smashing was detected during raminit when not loading from MRC.
Adding CAR_GLOBAL to a struct inside raminit was suggested in
https://mail.coreboot.org/pipermail/coreboot/2018-May/086677.html in
order to fix the problem.
Adding CAR_GLOBAL to the ram timings variable solves the issue (adding
it to the ram_training or raminfo struct had no effect).
This is just a workaround and might need a proper fix in the future.
Tested on Lenovo X201i with 2+2 and 4+4 GB RAM.
Change-Id: I21b380db61be2aedc045201821d83e18e7d07ad1
Signed-off-by: Matthias Gazzari <mail@qtux.eu>
Reviewed-on: https://review.coreboot.org/26388
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
The common mrc cache driver allows to save the raminit training
results to a separate fmap region which is more manageable than a
cbfsfile.
Change-Id: I25a6d3fe5466d142e3d10429a87b19047040c251
Signed-off-by: Arthur Heymans <arthur@aheymans.xyz>
Reviewed-on: https://review.coreboot.org/23484
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Patrick Rudolph <siro@das-labor.org>
Provide a valid ACPI path for coreboot's SSDT generators.
Fixes all ACPI errors found while booting GNU Linux 4.15 on
Lenovo T410.
Change-Id: Idd4986f39f21cb53cb019d0893d40fed94c6505b
Signed-off-by: Patrick Rudolph <siro@das-labor.org>
Reviewed-on: https://review.coreboot.org/26287
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Arthur Heymans <arthur@aheymans.xyz>
To calculate the CBMEM address we need to determine the framebuffer
size early in the ROMSTAGE. We now do the calculation before
cbmem_recovery() and configure the memory controller right away.
If the calculation was done from cbmem_top() instead, we'd loose some
logging that seems useful, since printk() would recurse to cbmem_top() too
with CONSOLE_CBMEM enabled.
If we didn't configure the memory controller at this point, we'd
need to store the result somewhere else. However, CAR_GLOBAL is not
practical at this point, because calling car_get_var() from cbmem_top()
would recurse back to cbmem_top().
Change-Id: Ib9ae0f97f9f769a20a610f8d76f14165fb924042
Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>
Reviewed-on: https://review.coreboot.org/25798
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
This memory controller supports both DDR2 and DDR3 memory, yet many
functions have ddr2 in their name while not being ddr2 specific.
This patch renames those to avoid confusion.
Change-Id: Ib3d10014f530905155e56fc52706edb4ab9f5630
Signed-off-by: Arthur Heymans <arthur@aheymans.xyz>
Reviewed-on: https://review.coreboot.org/19870
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
Some things in programming registers related to dual channel
interleaved operation were wrong.
This also adds some code that could in the future be used when me is
active and claims some memory for its UMA.
This also uses some more sensible variable names to clarify at least
some of the magic.
This fixes memtest86+ failing with some assymetric DIMM configuration.
TESTED on DG43GT: memtest86+ now succeeds on many more different DIMM
configuration setups (would instantly fail at addresses above 4G on
many configurations).
Change-Id: If84099d27100e57437bf214dc4cf975f67c2ea1f
Signed-off-by: Arthur Heymans <arthur@aheymans.xyz>
Reviewed-on: https://review.coreboot.org/22914
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
Add a function domain_acpi_name to return "PCI0", rather than
falling back to the parent' device's "\_SB" label. This repair is
required for the LPC TPM device to register its presence without
blowing up the table and preventing the payload from finding SATA.
Before change, the TPM device reported as:
\_SB.\_SB.LPC0.TPM
After change, the TPM device reports as:
\_SB.PCI0.LPC0.TPM
A separate change submission will correct "LPC0" as well.
Change-Id: I5e8d4715c9b42f50c84dd65818e4b0fdfc9d54f9
Signed-off-by: Kevin Cody-Little <kcodyjr@gmail.com>
Reviewed-on: https://review.coreboot.org/26204
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Nico Huber <nico.h@gmx.de>
The existing logic to set up CsMux67 used an incorrect mask
and comparison value due to a copy + paste editing error.
Use the correct mask and comparison value for the last two
values.
Commit cf1cb5b2d4 did the same
for CsMux45 but missed this one.
Change-Id: Ib97ca89535b8291397d42eca69e217c21a9dd937
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Reviewed-on: https://review.coreboot.org/25994
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
This change also makes sure that the sum the uma regions (TSEG, GSM,
GSM) is 4MiB aligned. This is needed to avoid cbmem_top floating between
2 usable ram region, since cbmem_top is aligned 4MiB down to easy MTRR
setup for ramstage. At least tianocore requires this and fails to boot
without it.
Better MTRR are achieved by making the memory 'hole' till 4GiB exactly
2Gib.
This code mimics how it is done in nb/intel/gm45 and achieves similar
results.
TSEG is enabled and set to 8M since this makes it easier to reuse the
common smm setup / parallel mp code and makes it possible to cache the
ramstage in there like how it's done on newer targets.
TESTED on Intel DG43GT.
Change-Id: I1b5ea04d9b7d5494a30aa7156d8c17170e77b8ad
Signed-off-by: Arthur Heymans <arthur@aheymans.xyz>
Reviewed-on: https://review.coreboot.org/21634
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Nico Huber <nico.h@gmx.de>
Programming CxDRB should be cumulative as explained in "Intel ® 4
Series Chipset Family datasheet".
This does not seem to have any real impact but better do according to
the documentation and what vendor firmware does.
This also removes some dead code.
Change-Id: I7ff3264824c843f84b9eb6c06a06aa3f151fe4b3
Signed-off-by: Arthur Heymans <arthur@aheymans.xyz>
Reviewed-on: https://review.coreboot.org/22911
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
This training find the optimal write DQ delay and read DQS delay
settings. It does so on all lanes at the same time, like
vendor (training each lane individually has poor results).
The results are stored in the sysinfo struct and restored on next
boots and S3 resume.
This potentially increases stability as optimal settings are chosen
and is more necessary for DDR3 raminit where the write DQS delays are
leveled/variable due to the flyby topology.
TESTED on Intel DG43GT with (2G + 1G) on each channel, see that the
results are quite close to the safe original ones (that previous
worked fine) and tested with memtest86+.
Change-Id: Iacdc63b91b4705d1a80437314bfe55385ea5b6c1
Signed-off-by: Arthur Heymans <arthur@aheymans.xyz>
Reviewed-on: https://review.coreboot.org/22329
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
Remove locking of PCI device 00:00.0 registers (nehalem/finalize.c)
and remove setting the zeroth bit of the MSR_LT_LOCK_MEMORY = 0x2e7 MSR
register (model_2065x/finalize.c) to fix a frozen boot and S3 resume issue
which became apparent with commit d533b16669.
More detailed, either setting the LSB of the 32 bit register at 0x98
of the PCI device 00:00.0 (in the intel_nehalem_finalize_smm function) or
setting the LSB of the the MSR register MSR_LT_LOCK_MEMORY = 0x2e7 (in the
intel_model_2065x_finalize_smm function) indepentenly causes a freeze
during bootup or a complete session loss on resuming from S3 as described
here: https://mail.coreboot.org/pipermail/coreboot/2018-April/086564.html
It seems like Nehalem CPUs do not have a MSR_LT_LOCK_MEMORY register.
Additionally, the "Intel Core i7-600, i5-500, i5-400 and i3-300 Mobile
Processor Series, Datasheet Volume Two" indicates that registers of the
PCI device 00:00.0 cannot be locked manually. Instead, they can only be
locked by TXT, VT-d, CMD.LOCK.MEMCONFIG, ME_SM_LOCK or D_LCK.
Finally, the addresses and sizes of these registers were partially wrong.
Tested on Lenovo X201i with a Core i3 330M (no AES-NI, no VT-d and no TXT
support compared to the Core i5 and Core i7 processors of a X201).
Change-Id: I9d568d5c05807ebf7e131b3e5be8e5445476d61b
Signed-off-by: Matthias Gazzari <mail@qtux.eu>
Reviewed-on: https://review.coreboot.org/25914
Reviewed-by: Nicola Corna <nicola@corna.info>
Reviewed-by: Jonathan Neuschäfer <j.neuschaefer@gmx.net>
Reviewed-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-by: Nico Huber <nico.h@gmx.de>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
