Add a bootblock.c file for the northbridge and setup the
PCIEXBAR as the first thing using IO PCI config acceses.
After that all PCI config accesses can use MMIO.
Change-Id: I51d229c626c45705dda1757c2f14265cbc0e6183
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/2617
Tested-by: build bot (Jenkins)
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
The Haswell parts use a PCH code named Lynx Point (Series 8). Therefore,
the southbridge support is included as well. The basis for this code is
the Sandybridge code. Management Engine, IRQ routing, and ACPI still requires
more attention, but this is a good starting point.
This code partially gets up through the romstage just before training
memory on a Haswell reference board.
Change-Id: If572d6c21ca051b486b82a924ca0ffe05c4d0ad4
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/2616
Tested-by: build bot (Jenkins)
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
The original code attempted to reserve a space in RAM for coreboot to
remain resident. This turns out not to be needed, and breaks things
for the kernel since the exynos5250-smdk5250 kernel device tree starts
RAM at 0x40000000.
(This patch was originally by Gabe, I'm just uploading it)
Change-Id: I4536edaf8785d81a3ea008216a2d57549ce5edfb
Signed-off-by: Gabe Black <gabeblack@chromium.org>
Signed-off-by: David Hendricks <dhendrix@chromium.org>
Reviewed-on: http://review.coreboot.org/2698
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
Tested-by: build bot (Jenkins)
Change the OSC method to actually grant control of
PCIe capabilities to the OS instead of granting no
control. I believe the logic was backwards in the
original commit. Bits should be set when granting
control and cleared when not granting control. By
setting the return value to 0x00, we effectively
tell the OS that it cannot control any PCIe
capability. See section 6.2.9 of the ACPI spec
version 3.0 for more information.
This edit is a duplication of the OSC method that
is in the src/southbridge/intel/bd82x6x/pch.asl
file.
Change-Id: Id2462ab12203afceb9033f24d06b4dfbf2236d2e
Signed-off-by: Mike Loptien <mike.loptien@se-eng.com>
Reviewed-on: http://review.coreboot.org/2714
Tested-by: build bot (Jenkins)
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
This enables branch prediction. We can probably find a better place
to do this, but for now we'll do it in snow's romstage main().
Change-Id: I86c7b6bc9e897a7a432c490fb96a126e81b8ce72
Signed-off-by: David Hendricks <dhendrix@chromium.org>
Reviewed-on: http://review.coreboot.org/2701
Tested-by: build bot (Jenkins)
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
Since commit »Drop redundant CHIP_NAME in mainboard.c« (a93c3fe7) [1]
`CHIP_NAME` is unneeded for mainboards as the name is composed
automatically in `src/devices/root_device.c` from the strings in
Kconfig.
Unfortunately the ports for Google Butterfly, Link and Parrot as
as well as IEI PM-LX2-800-R10 introduced CHIP_NAME again. So drop
it again too.
[1] http://review.coreboot.org/1635
Change-Id: Ice7577a2a5c6070e196f2647c440b7a8e140e27e
Signed-off-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-on: http://review.coreboot.org/2708
Tested-by: build bot (Jenkins)
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
PS_HOLD gets set in exynos' power_init().
Change-Id: Ib08e0afcad23cbd07dc7e3727fd958a1bc868b5a
Signed-off-by: David Hendricks <dhendrix@chromium.org>
Reviewed-on: http://review.coreboot.org/2700
Reviewed-by: Hung-Te Lin <hungte@chromium.org>
Tested-by: build bot (Jenkins)
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
Call the power_init() function. We appear to have forgotten about it
when deprecating lowlevel_init_subsystems(), but it didn't seem to
cause problems until we got to doing more interesting stuff recently.
There are some clean-ups to do from the original code, such as not
attempting to configure I2C from PMIC code, which we'll get around
to in follow-up patches.
(Credit to Gabe for spotting this)
Change-Id: I6a59379e9323277d0b61469de9abe6d651ac5bfb
Signed-off-by: Gabe Black <gabeblack@chromium.org>
Signed-off-by: David Hendricks <dhendrix@chromium.org>
Reviewed-on: http://review.coreboot.org/2699
Reviewed-by: Hung-Te Lin <hungte@chromium.org>
Tested-by: build bot (Jenkins)
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
The current default is IDE mode which is slower compared to AHCI
mode. Therefore use AHCI mode by default.
A similar change was made for AMD Persimmon in commit
»Enable SATA AHCI for faster boot with SeaBIOS.« (96be74c7) [1]
but was indirectly reverted by »sb800: Add sata ahci/raid mode
kconfig option« (d4a0e7d0) [2].
[1] http://review.coreboot.org/220
[2] http://review.coreboot.org/225
Change-Id: I4fa31b0a3280891e7a3f37675ae8415205818947
Signed-off-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-on: http://review.coreboot.org/2661
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marc.jones@se-eng.com>
There's a compile time error that we didn't catch since the
board defaults as used by the build bot won't expose it.
Just make watchdog_off() a no-op statement so there aren't any
stray semicolons in the preprocessor output.
Change-Id: Ib5595e7e8aa91ca54bc8ca30a39b72875c961464
Reported-by: 'lautriv' on irc.freenode.net/#coreboot
Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Reviewed-on: http://review.coreboot.org/2627
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
We used to allow mainboards to override subsystems using
mainboard_pci_subsystem_vendor_id and mainboard_pci_subsystem_device_id.
Mechanisms have changed and the only occurrence of these names is in
the header.
Change-Id: Ic2ab13201a2740c98868fdf580140b7758b62263
Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Reviewed-on: http://review.coreboot.org/2625
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Tested-by: build bot (Jenkins)
Despite everywhere the model name M5A88-V is used, in Kconfig the
string M5A88PM-V is used. Searching for that model string on the
WWW does not return anything which is unrelated to coreboot, so
change that string to M5A88-V.
Change-Id: I25cf9d4a5fc3f9b9356e8616452066ebf873f44c
Signed-off-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-on: http://review.coreboot.org/2613
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martin.roth@se-eng.com>
Reviewed-by: QingPei Wang <wangqingpei@gmail.com>
Add PEI updates and ACPI updates for supporting EHCI to XHCI
USB port support.
Change-Id: I9ace68a1b3950771aefb96c1319b8899291edd9a
Signed-off-by: Marc Jones <marc.jones@se-eng.com>
Reviewed-on: http://review.coreboot.org/2519
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martin.roth@se-eng.com>
Adding the 'WordBusNumber' macro to the PCI0
CRES ResourceTemplate in the Persimmon DSDT.
This sets up the bus number for the PCI0 device
and the secondary bus number in the CRS method.
This change came in response to a 'dmesg' error
which states:
'[FIRMWARE BUG]: ACPI: no secondary bus range in _CRS'
By adding the 'WordBusNumber' macro, ACPI can set
up a valid range for the PCIe downstream busses,
thereby relieving the Linux kernel from "guessing"
the valid range based off _BBN or assuming [0-0xFF].
The Linux kernel code that checks this bus range is
in `drivers/acpi/pci_root.c`. PCI busses can have
up to 256 secondary busses connected to them via
a PCI-PCI bridge. However, these busses do not
have to be sequentially numbered, so leaving out a
section of the range (eg. allowing [0-0x7F]) will
unnecessarily restrict the downstream busses.
This change will apply to other AMD mainboards and
will be in a different commit.
Change-Id: I44f22bc03a0dcbcd2594d4291508826cc2146860
Signed-off-by: Mike Loptien <mike.loptien@se-eng.com>
Reviewed-on: http://review.coreboot.org/2592
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martin.roth@se-eng.com>
Reviewed-by: Marc Jones <marc.jones@se-eng.com>
This eliminates the use of do_div() in favor of using libgcc
functions.
This was tested by building and booting on Google Snow (ARMv7)
and Qemu (x86). printk()s which use division in vtxprintf() look good.
Change-Id: Icad001d84a3c05bfbf77098f3d644816280b4a4d
Signed-off-by: Gabe Black <gabeblack@chromium.org>
Signed-off-by: David Hendricks <dhendrix@chromium.org>
Reviewed-on: http://review.coreboot.org/2606
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
Changes:
- Get rid of the inagua mainboard specific code and use the
platform generic function wrapper that was added in change
http://review.coreboot.org/#/c/2497/
AMD f14: Add SPD read functions to wrapper code
- Move DIMM addresses into devicetree.cb
- Add the ASF init that used to be in the SPD read code into
mainboard_enable()
Notes:
- The DIMM reads only happen in romstage, so the function is not
available in ramstage. Point the read-SPD callback to a generic
function in ramstage.
Change-Id: Id05227fcf18c6ab94ffe1beb50b533ab7b0535db
Signed-off-by: Kimarie Hoot <kimarie.hoot@se-eng.com>
Reviewed-on: http://review.coreboot.org/2607
Reviewed-by: Martin Roth <martin.roth@se-eng.com>
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Currently for Advansus A785E-I, ASRock E350M1 and ASUS M5A88-V
despite what is chosen in Kconfig »Chipset« menu item,
$ more .config
[…]
# CONFIG_ENABLE_IDE_COMBINED_MODE is not set
CONFIG_IDE_COMBINED_MODE=0x1
# CONFIG_SB800_SATA_IDE is not set
CONFIG_SB800_SATA_AHCI=y
# CONFIG_SB800_SATA_RAID is not set
CONFIG_SB800_SATA_MODE=0x2
[…]
the SATA controller is put into IDE mode.
$ lspci -nn | grep SATA
00:11.0 SATA controller [0106]: Advanced Micro Devices [AMD] nee ATI SB7x0/SB8x0/SB9x0 SATA Controller [IDE mode] [1002:4390] (rev 40)
Commit »sb800: Add sata ahci/raid mode kconfig option«
(d4a0e7d0) [1] added the options above to configure the mode
using Kconfig and some SB800 boards were adapted already. For
example commit »persimmon: sb800 sata mode configure update«
(1386fa74) [2] did so for AMD Persimmon.
Doing the same by assigning the Kconfig variable to the value in
`platform_cfg.h` integrates this with the three remaining boards
listed above.
The patch is successfully tested with the ASRock E350M1.
$ lspci -nn | grep SATA
00:11.0 SATA controller [0106]: Advanced Micro Devices [AMD] nee ATI SB7x0/SB8x0/SB9x0 SATA Controller [AHCI mode] [1002:4391] (rev 40)
[1] http://review.coreboot.org/225
[2] http://review.coreboot.org/227
Change-Id: I227257e2c8f04f18c27ff00fe62d42e372de67e4
Signed-off-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-on: http://review.coreboot.org/2610
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martin.roth@se-eng.com>
Replace »persimmon« by »board« in comment to keep `diff` output
between boards small.
Change-Id: Ieae2a63782c488ae35f22eb30f5b1049200d12c8
Signed-off-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-on: http://review.coreboot.org/2611
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martin.roth@se-eng.com>
Changes:
- Get rid of the union_station mainboard specific code and
use the platform generic function wrapper that was added
in change http://review.coreboot.org/#/c/2497/
AMD f14: Add SPD read functions to wrapper code
- Move DIMM addresses into devicetree.cb
- Add the ASF init that used to be in the SPD read code into
mainboard_enable()
Notes:
- The DIMM reads only happen in romstage, so the function is not
available in ramstage. Point the read-SPD callback to a generic
function in ramstage.
Change-Id: I19d6b0d674b67294519383f80928471b37da1e14
Signed-off-by: Kimarie Hoot <kimarie.hoot@se-eng.com>
Reviewed-on: http://review.coreboot.org/2609
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martin.roth@se-eng.com>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Changes:
- Get rid of the south_station mainboard specific code and
use the platform generic function wrapper that was added
in change http://review.coreboot.org/#/c/2497/
AMD f14: Add SPD read functions to wrapper code
- Move DIMM addresses into devicetree.cb
- Add the ASF init that used to be in the SPD read code into
mainboard_enable()
Notes:
- The DIMM reads only happen in romstage, so the function is not
available in ramstage. Point the read-SPD callback to a generic
function in ramstage.
Change-Id: If4291d25ea81bf375f55b64c07c223a847a211d0
Signed-off-by: Kimarie Hoot <kimarie.hoot@se-eng.com>
Reviewed-on: http://review.coreboot.org/2608
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martin.roth@se-eng.com>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
This is previously used exception code from libpayload.
On startup it installs and then tests an exception handler.
The test is an unaligned memory operation.
Yes, we've seen what might be exceptions in the ramstage, and
it makes sense to handle them. This code is identical in structure
and operation to the previously committed payload exception handler,
though we reserve the right to change it as circumstances require.
The remaining question is whether we need it in romstage.
Change-Id: I24484686c33c9757af8ba171ebae9773828fb69d
Signed-off-by: Gabe Black <gabeblack@google.com>
Signed-off-by: Ronald G. Minnich <rminnich@gmail.com>
Reviewed-on: http://review.coreboot.org/2614
Tested-by: build bot (Jenkins)
Reviewed-by: David Hendricks <dhendrix@chromium.org>
AGESA code has wrong definition of CR0_PE bit (1 instead of 0).
PE [Protected Mode Enable] is 0 bit in CR0 register
(If PE=1, system is in protected mode, else system is in real mode)
Bit 1 is MP [Monitor co-processor]
(Controls interaction of WAIT/FWAIT instructions with TS flag in CR0)
System uses CR0_PE define, but I didn't expect any consequences because of this bug.
Change-Id: I54d9a8c0ee3af0a2e0267777036f227a9e05f3e1
Signed-off-by: Konstantin Aladyshev <aladyshev@nicevt.ru>
Reviewed-on: http://review.coreboot.org/2591
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marc.jones@se-eng.com>
According to BKDG:
"Memory controller (MCT) and DRAM controllers (DCTs) additions:
• Support for 933 MHz (1866 MT/s) MEMCLK frequency."
Change-Id: I6f307ce3fcb355d5445f1ea86def73a41b928a57
Signed-off-by: Konstantin Aladyshev <aladyshev@nicevt.ru>
Reviewed-on: http://review.coreboot.org/2589
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marc.jones@se-eng.com>
_RDMSR instruction loads the contents of a 64-bit model specific register (MSR)
specified in the ECX register into registers EDX:EAX.
The EDX register is loaded with the high-order 32 bits of the MSR
and the EAX register is loaded with the low-order 32 bits.
EDX:EAX = MSR[ECX]
So bit 49 will be contained in EDX register.
Buggy code instead of bit 49 (CombineCr0Cd) sets bit [49-32=17] (PfcStrideDis).
PfcStrideDis bit disables stride prefetch generation. This leads to memory
bandwidth loss.
_________
Supermicro H8QGI board
After applying this change i observed huge memory bandwidth increase in tests
that runs on small amount of cores. But unfortunately it doesn't affect
overall bandwidth results on 4P system with 48 cores.
So i think that in this system leading limiting factor is
AMD HT-ASSIST feature (Probe filter).
But right now it is not working. System stucks in Linux boot. I have done
some experiments and figured out that stuck happens when system have cores in
compute unit (CU) other than CU with BSC (boot strap core).
CU is two cores (primary and seconary) that shares some things (L2 cache, FPU ...)
So with probe filter i can boot Linux with one (BSC)
or two (BSC + secondary core in its CU) cores.
And with this configuration i can see memory bandwidth on 1 core (or two cores)
close to original bios.
Change-Id: I5a95f5b753d600c70d3c93d36fecc687610c61cd
Signed-off-by: Konstantin Aladyshev <aladyshev@nicevt.ru>
Reviewed-on: http://review.coreboot.org/2588
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marc.jones@se-eng.com>
The Hudson-E1's default SPI speed for normal i.e. non-fast reads is 66 MHz,
but the SST 25VF032B datasheet allows max. 25. Lower the speed to 22 MHz,
otherwise BIOS flashing fails.
Change-Id: I22e87d833a3ebd316b6e873595a2480831533ab1
Signed-off-by: Jens Rottmann <JRottmann@LiPPERTembedded.de>
Reviewed-on: http://review.coreboot.org/2605
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Changes:
- Get rid of the persimmon mainboard specific code which has been
moved into the wrapper as a platform generic function in change
http://review.coreboot.org/#/c/2497/
AMD f14: Add SPD read functions to wrapper code
- Move DIMM addresses into devicetree.cb
- Add the ASF init that used to be in the SPD read code into
mainboard_enable()
Notes:
- The DIMM reads only happen in romstage, so the function is not
available in ramstage. Point the read-SPD callback to a generic
function in ramstage.
Change-Id: I5f017dbb8dee5a09ec19734a6069ff9b71a6ab50
Signed-off-by: Martin Roth <martin.roth@se-eng.com>
Reviewed-on: http://review.coreboot.org/2500
Tested-by: build bot (Jenkins)
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
Reviewed-by: Jens Rottmann <JRottmann@LiPPERTembedded.de>
Reviewed-by: Marc Jones <marc.jones@se-eng.com>
Change:
This is the initial step for moving the AMD F14 & HUDSON1,2,3
SPD-read callout out of the mainboard directories and into
the wrapper. The next step is to update the platforms to use
this routine in BiosCallouts.c and to delete the code from the
mainboard directories. The DIMM addresses should be moved into
devicetree.cb.
If there are significant differences or reasons that the mainboard
needs to override this code, it's perfectly reasonable to keep using
the version in the mainboard, but this allows us to remove duplicated
code and simplify the mainboard directories.
Notes:
This started by duplicating what was in Persimmon, and was changed to
use the devicetree.cb structures. The ASF setup was also removed from
the persimmon copy (PMIO writes to 0x28 & 0x29) as that's not needed
for the SPD access and doesn't make sense to initialize here.
Significant cleanup and magic number reduction was done as well.
It is intended that this file will not be included in ramstage as
the DIMM init is all done in romstage.
This is similar to what was done for Parmer/Thatcher in commit
7fb692bd - http://review.coreboot.org/#/c/2190/
Fam15tn: Move SPD read from mainboards into wrapper
Yes, it would make sense to split this into two separate files
and move the SMBUS initialization and access into the southbridge
wrapper. Maybe that can come next.
Change-Id: I1e106d3912c160b0015bf02158d9faba4f578ee3
Signed-off-by: Martin Roth <martin.roth@se-eng.com>
Reviewed-on: http://review.coreboot.org/2497
Tested-by: build bot (Jenkins)
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
Reviewed-by: Jens Rottmann <JRottmann@LiPPERTembedded.de>
Reviewed-by: Marc Jones <marc.jones@se-eng.com>
I've used an operating system for over 10 years now that makes
UTF-8 easy. It's not called Linux or OSX.
When UTF-8 is needed, of course, then we can look again.
I can't think of a single redeeming feature of placing
it in the comment in this manner. It's certainy not
needed.
The inclusion of UTF-8 characters is inconvenient,
especially from a text terminal.
I don't really want to start using compose in
CROSH shell terminals on chromeos.
We might want to incorporate "no UTF-8" as a
commit filter. For now, get rid of these
characters.
Change-Id: If94cc657bae1dbd282bec8de6c5309b1f8da5659
Signed-off-by: Ronald G. Minnich <rminnich@gmail.com>
Reviewed-on: http://review.coreboot.org/2604
Reviewed-by: Bernhard Urban <lewurm@gmail.com>
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
This reverts commit 1cd6160821
Division bites us again. I don't know how or why, but printk() seems to break (again) with this patch. I'm surprised we didn't encounter problems earlier on...
Change-Id: I81cb9f20879f5eb73a76e1af47b96a68d1e81dc8
TODO: Find a better solution for div64. This one is too painful, but seems necessary for now (and sort-of works with our vtxprintf hack).
Reviewed-on: http://review.coreboot.org/2600
Tested-by: build bot (Jenkins)
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
This adds some real GPIO mappings where virtual GPIOs were used before.
Change-Id: I25d4be45f986c8d622b97151f8bdae2651baf3e6
Signed-off-by: David Hendricks <dhendrix@chromium.org>
Reviewed-on: http://review.coreboot.org/2603
Tested-by: build bot (Jenkins)
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
This adds an enum for GPIO ports on the Exynos5. To make them
useful, they are assigned the absolute MMIO address where a
s5p_gpio_bank struct can point to.
Change-Id: Ia539ba52d7393501d434ba8fecde01da37b0d8aa
Signed-off-by: David Hendricks <dhendrix@chromium.org>
Reviewed-on: http://review.coreboot.org/2602
Tested-by: build bot (Jenkins)
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
While reading through the file fix some spotted errors like
indentation, locution(?), capitalization and missing full stops.
Change-Id: Id435b4750e329b06a9b36c1df2c39d2038a09b18
Signed-off-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-on: http://review.coreboot.org/2484
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martin.roth@se-eng.com>
Reviewed-by: Marc Jones <marc.jones@se-eng.com>
A board without HAVE_ACPI_RESUME did not build with
COLLECT_TIMESTAMPS enabled as `cbmem.c` was not built.
Change-Id: I9c8b575d445ac566a2ec533d73080bcccc3dfbca
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: http://review.coreboot.org/2549
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marc.jones@se-eng.com>
Quoting Jens Rottmann [1]:
Nevertheless I still think this whole function is bogus for the E350M1. The
function assumes GPIO21 is wired to reset APU PCIe lane 0+1 (PCIe x8, port 4+5
as Coreboot/AGESA calls it), GPIO25 resets lane 2 (PCIe x4) and GPIO02 lane 3.
But the E350M1 has PCIe x16 i.e. probably APU lanes 0-3 bundled, completely
different layout. They could have chosen GPIO21 to force resets, or 25 - or
maybe 50 like on the Persimmon or any other they fancied or - and this is the
most probable - none at all. Having BiosGnbPcieSlotReset() toggle some GPIOs
without knowing what they do on the E350M1 (if anything at all) is nonsense.
In my opinion this whole function should just "return AGESA_UNSUPPORTED" and
good riddance.
[1] http://review.coreboot.org/#/c/2445/
Change-Id: Iac66da41182e838c7e6925250cc3982adbb3e4ec
Reported-by: Jens Rottmann <JRottmann@LiPPERTembedded.de>
Signed-off-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-on: http://review.coreboot.org/2489
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marc.jones@se-eng.com>
Reviewed-by: Jens Rottmann <JRottmann@LiPPERTembedded.de>
These are essential functions for setting up the display port and
framebuffer, and also enable such things as aux channel
communications. We do some very simple initialization in romstage,
mainly set a GPIO so that the graphics is powering up, but the complex
parts are done in the ramstage. This mirrors the way in which graphics
is done in the x86 size.
I've added a first pass at a real device, and put it in the mainboard
Kconfig, hoping for corrections. Because startup is so complex,
depending on device type, I've created a 'displayport' device that
removes some of the complexity and makes the flow *much* clearer. You
can actually follow the flow by looking at the code, which is not true
on other implementations. Since display port is perhaps the main port
used on these chips, that's a reasonable compromise. All parameters of
importance are now in the device tree.
Change-Id: I56400ec9016ecb8716ec5a5dae41fdfbfff4817a
Signed-off-by: Ronald G. Minnich <rminnich@gmail.com>
Reviewed-on: http://review.coreboot.org/2570
Tested-by: build bot (Jenkins)
Since the merg of the ASRock E350M1 port (a649a96e) the compiler
warns about the following [1].
mainboard.c:35, GNU Compiler 4 (gcc), Priorität: Normal
no previous prototype for 'set_pcie_reset' [-Wmissing-prototypes]
mainboard.c:43, GNU Compiler 4 (gcc), Priorität: Normal
no previous prototype for 'set_pcie_dereset' [-Wmissing-prototypes]
Adding the function prototypes to the beginning of the file as
done in commit »Persimmon updates for AMD F14 rev C0« (d7a696d0)
addresses the warning.
[1] http://qa.coreboot.org/job/coreboot-gerrit/4975/warnings13Result/package.-139448264/file.-1544928473/
[2] http://review.coreboot.org/137
Change-Id: Iad2e62ec37c3a2f749a264974b61ac7c226e9b83
Signed-off-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-on: http://review.coreboot.org/2590
Tested-by: build bot (Jenkins)
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
Set up the clocks used for sound and turn on the sound clock.
Change-Id: Ic59bfa9ae87116299503e6d25aeefba98c842fb8
Signed-off-by: Gabe Black <gabeblack@google.com>
Signed-off-by: Ronald G. Minnich <rminnich@gmail.com>
Reviewed-on: http://review.coreboot.org/2587
Tested-by: build bot (Jenkins)
Reviewed-by: David Hendricks <dhendrix@chromium.org>
The MMC0 on google/snow can run in 8 bit mode. To simplify driver development,
we thought disabling it (using zero, which runs in 1-bit / 4-bit mode) may help.
However, after some experiments in payload drivers, setting pinmux to 8 bit mode
can still allow MMC to run in 1-bit / 4-bit mode, so it's pretty safe to enable
8 bit mode by default for better performance.
Verified to boot on google/snow, and got MMC0 working.
Change-Id: Ic0acc723fe6a8aecf373429d3801beadd70815d9
Signed-off-by: Ronald G. Minnich <rminnich@gmail.com>
Reviewed-on: http://review.coreboot.org/2585
Tested-by: build bot (Jenkins)
Reviewed-by: David Hendricks <dhendrix@chromium.org>
The power up default for the 14M_25M_48M_OSC switchable clock output ball of
the SB800 chipset is 14 MHz. sb800/bootblock.c changes this to 48 MHz,
which is the correct value for almost all SIOs. However, not for
'smscsuperio' (SMSC SCH311x), which needs the original 14 MHz and is not
configurable for other clock speeds. A wrong SIO clock supply results in
funny RS232 output (wrong bit speed) and non-working PS/2.
We could switch back to 14 MHz in the mainboard's romstage.c, but then the
clock frequency would change twice. The resulting short 48 MHz burst causes
a handful of rubbish characters on RS232 on every boot until the SIO clock
has stabilized again.
This patch skips the SB800 clock switch if the SIO Kconfig requests 14 MHz.
This does not affect any boards currently in the repository (yet).
Change-Id: Icff41fd88dc41c08f3700ab4f786852f04eff2a4
Signed-off-by: Jens Rottmann <JRottmann@LiPPERTembedded.de>
Reviewed-on: http://review.coreboot.org/2454
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Martin Roth <martin.roth@se-eng.com>
The first reason for selecting the CPU model at compile time was a
multi-second pause if booting a single core Fusion T40R with MAX_CPUS=2.
Recent tests show the pause has disappeared, someone must have fixed it.
The second reason was me not knowing how to make a single vgabios image
work with two different PCI IDs. Many thanks to Martin Roth for educating
me! Quote:
"The way to make coreboot use the same vbios for different video device IDs
is through the map_oprom_vendev function. In family 14 it's in
northbridge/amd/agesa/family14/amdfam14_conf.c You would name your video
bios 1002,9802 in the config and all the other device/vendor IDs for the
family 14h processors will fall through the initial check for the video
bios and will get remapped to use that vbios. This only works if you're
initializing the vbios inside coreboot. I don't know if you're using
SeaBios as a payload, but if you are you can add the vbios to cbfs as
vgaroms/vbios.rom and the rom will always be initialized."
I'd like to add the vgabios is added as type 'optionrom' when Coreboot make
adds it, however to work with SeaBios it has to be added manually with
cbfstool and with type 'raw', or it will hang.
Change-Id: I8190d0c3202a60dfccb77dde232f9ba7ce5ce318
Signed-off-by: Jens Rottmann <JRottmann@LiPPERTembedded.de>
Reviewed-on: http://review.coreboot.org/2584
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
OK, this is tl;dr. But I need to write this in hopes we make
sure we don't put code like this into coreboot. Ever.
Our excuse in this case is that it was imported, not obviously wrong,
and easily changed. It made sense to get it in, make it work, then
do a cleanup pass, because changing everything up front is almost
impossible to debug.
The exynos code has bunch of base register values, e.g.
These are base addresses of things that look like a memory-mapped
struct. To get these to a pointer, they created the following macro,
which creates an inline function.
static inline unsigned int samsung_get_base_##device(void) \
{ \
return cpu_is_exynos5() ? EXYNOS5_##base : 0; \
}
And then invoke it 31 times in a .h file, e.g.:
SAMSUNG_BASE(clock, CLOCK_BASE)
to create 31 functions.
And then use it:
struct exynos5_clock *clk =
(struct exynos5_clock *)samsung_get_base_clock();
OK, what's wrong with this? It's easier to ask what's right with it. Answer: nothing.
I have a long list of what's wrong, and I may leave some things out,
but here goes:
1. the "function" can return a NULL if we're not on exynos5. Most uses of the code
don't check the return value.
2. And why would this function be running, if we're not on an exynos5? Why compile it in?
3. Note the cast everywhere a samsung_get_base_xxx is used.
The function returns an untyped variable, requiring the *user* to get two
things right: the cast, and the function invocation. One can replace that _clock(); with
_power(); in the code above, and they will be referencing the wrong registers, and
they'll never get an error!
We have a C compiler; use it to type data.
4. You're generating 31 functions using cpp each and every time the file is included.
The C compiler has to parse these each time. It's not at all like a simple cpp
macro which is only generated on use.
5. You can't tags or etags this code
6. In fact, any kind of analysis tool will be unable to do anything with this cpp magic.
That's only a partial list.
So what's the right way to do it? Just make typed constants, viz:
Or, since I expect people will want the lower case function syntax, I've left
it that way:
Now we've got something that is efficient, and we don't even need to protect with
any more.
Hence this change. We've got something that is type checked, does not require users to
cast on each use, will catch simple programming errors, can be analyzed with standard tools,
and builds faster.
So if we make a mistake:
struct exynos5_clock *clk =
samsung_get_base_adc();
We'll see it:
src/cpu/samsung/exynos5250/clock.c: In function 'get_pll_clk':
src/cpu/samsung/exynos5250/clock.c:183:3: error: initialization from incompatible pointer type [-Werror]
which we would not have seen before.
As a minor benefit, it shaves most of a second off the compilation.
Change-Id: Ie67bc4bc038a8dd1837b977d07332d7d7fd6be1f
Signed-off-by: Ronald G. Minnich <rminnich@gmail.com>
Reviewed-on: http://review.coreboot.org/2582
Tested-by: build bot (Jenkins)
Update coreboot to use SeaBIOS' tag rel-1.7.2.1
Change-Id: I01969407964a7cf64f7c4800b59c6aed845b24f9
Signed-off-by: Idwer Vollering <vidwer@gmail.com>
Reviewed-on: http://review.coreboot.org/2575
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
Commit f154c018
Author: Marc Jones <marcj303@gmail.com>
Date: Wed Dec 14 11:24:00 2011 -0700
Persimmon audio codec verb patch.
Reviewed-on: http://review.coreboot.org/490
has a typo code*c* in the comments for `AZALIA_OEM_VERB_TABLE`. As
this was copied over to the LiPPERT Fam14 boards, use the following
command to fix the typo.
$ git grep -l cocec | xargs sed -i s,cocec,codec,
Change-Id: I1525b0445edab81ab136b3adece52b78ba7abc71
Signed-off-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-on: http://review.coreboot.org/2576
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martin.roth@se-eng.com>
If you try to reset the system with outb(3,0x92), outb(4,0xcf9) or a
triple-fault it will instead crash with a messy screen. As the more common
outb(0xFE, 0x64) doesn't work with our setup, Linux will crash whenever you
ask it to reboot. Closer inspection shows that on a warm boot of Coreboot
agesawrapper_amdinitpost() always fails with error code 7. Looks like DDR3
re-init goes wrong somehow. I tried find the reason for this but was
unable to. I am convinced this is not board specific but a bug in AGESA.
In the end I had to settle for a workaround: if amdinitpost returns 7 this
patch resets the system harder with outb(0x06, 0x0cf9), after that RAM init
will succeed. As amdinitpost is early in POST this automatic reset is
quick enough not to be noticable.
I'd perfer a real fix, but that's all I have.
Change-Id: I4763254b489f42a135232e45328ecf0d5c4d961a
Signed-off-by: Jens Rottmann <JRottmann@LiPPERTembedded.de>
Reviewed-on: http://review.coreboot.org/2573
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
Step 2: change the Persimmon code to adapt it to the new board's hardware.
The Toucan-AF is a COM Express Compact Type 6 form factor embedded board:
- AMD Fusion G-T56N (1.65 GHz dual core) or T40R (1 GHz single core) APU
- 1-4 GB DDR3 memory down
- 1x VGA, 2x DisplayPort (1 switchable to LVDS)
- AMD A55E (Hudson-E1) southbridge
- 8x USB 2.0
- 4x SATA
- HD Audio (with codec on baseboard)
- NEC uPD78F0532 microcontroller on I2C ("SEMA")
- 7x PCIe2.0 x1 (1 on PEG)
- Intel I210 GbE (on APU PCIe x1, can be disabled for additional PCIe)
- 2x SST 25VF032B (SO8, soldered) 4 MB SPI flash (BIOS and failsafe BIOS)
The Toucan-AF has no SIO on board. This patch includes basic support for a
Winbond W83627DHG (PS/2, 2x RS232), because the ADLINK ExpressBase-6 used
for evaluation happens to have one. The code may have to be adapted to the
actual baseboard of the application.
http://www.adlinktech.com/PD/web/PD_detail.php?pid=1132
Change-Id: I9041b905bad45852ac9b402fcbd5decbc98b377b
Signed-off-by: Jens Rottmann <JRottmann@LiPPERTembedded.de>
Reviewed-on: http://review.coreboot.org/2572
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
Step 1: copy all files unmodified from Persimmon. This makes it much
easier later to see how the two boards actually and deliberately differ
when porting bugfixes from one to the other. Git's copy detection is
imperfect (and slow).
Change-Id: I1ff02913479c07679f8c3ae5e6dd7876e6000b55
Signed-off-by: Jens Rottmann <JRottmann@LiPPERTembedded.de>
Reviewed-on: http://review.coreboot.org/2571
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
Aaron Durbin