When USB3 devices are attached while in suspend, or two USB3 devices
that are both plugged in are switched to the other port while in
suspend the kernel does not seem to notice this -- despite the cold
attach status bit. This results in the devices showing up in the USB
list at the old enumerated device numbers and higher layers continuing
to think they are present but not reseponding.
With the kernel workaround to deal with devices that are logically
disconnected it is possible for firmware to send a warm port reset to
devices that are in this state and then the kernel will see them disappear
and handle it properly.
This same issue exists in the EFI firmware on the Whitetip Mountain 2
reference board so it is not specifically a coreboot bug. If this
behavior is fixed in the kernel then this workaround could be removed
since it is in RW firmware.
BUG=chrome-os-partner:22818
BRANCH=falco,peppy,wolf,leon
TEST=manual:
1) attach two USB3 devices
2) suspend system
3) switch the ports that the USB3 devices are attatched to
4) resume system
5) confirm that the devices are re-enumerated and come up properly
Original-Change-Id: Ifba3ffc94a06dc0b2436d7d7d464d824657362af
Signed-off-by: Duncan Laurie <dlaurie@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/170335
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
(cherry picked from commit 203d200268f4af6445224962190cbc66ad2a83e4)
Change-Id: I54fd2847ee25a60f25c2cefebdc1a3c18455464a
Signed-off-by: Duncan Laurie <dlaurie@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/170579
[pm: rebase to master branch of coreboot upstream]
Signed-off-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-on: http://review.coreboot.org/6017
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
I have been attempting to work around USB3 issues that appear in the
kernel with hacks in the firmware, but this is resulting in more
headaches in the kernel.
Instead remove all the work that was being done at resume time and undo
the change that was issuing a warm reset to all ports at suspend time.
The bad device behavior will be dealt with at the kernel level to
handle devices that get stuck in polling state after enable/disable
sequence.
BUG=chrome-os-partner:22754
BRANCH=falco,peppy,wolf,leon
TEST=manual:
suspend/resume with several misbehaving devices:
Kingston USB3 Media Reader
Transcend USB3 Media Reader
Various ADATA USB3 drives
Various Kingston USB3 sticks
Original-Change-Id: I0894454af42d2ced456fe0da921d74c9e74902d0
Signed-off-by: Duncan Laurie <dlaurie@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/170107
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
(cherry picked from commit c2abb4d0dad6ed00e1e230d604c4c0a76eb4eef7)
Change-Id: Ib215d9c230f90a1c9f34bf29254bb9feec28c67e
Signed-off-by: Duncan Laurie <dlaurie@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/170578
[pm: rebase to master branch of coreboot upstream]
Signed-off-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-on: http://review.coreboot.org/6016
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
Worked out the purpose of more int15 calls and let them return
appropriate values. Also remove handlers for copy-pasted calls never
observed on this board.
Change-Id: I3d8c4ec5542bd19baca1dca83badc9b568779e1b
Signed-off-by: Nico Huber <nico.huber@secunet.com>
Reviewed-on: http://review.coreboot.org/6249
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
Fix some outputs of the super i/o that should be GPIOs and make
variables out of magic values that configure LVDS.
Change-Id: Ib9eef065980cefff0046485549a68cf8f070d5b9
Signed-off-by: Nico Huber <nico.huber@secunet.com>
Reviewed-on: http://review.coreboot.org/6248
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
The global pointer `biosmem` defined in vbe.c was never set. Thus, VBE
calls didn't work within YABEL.
Change-Id: I63c1c77755f9c442cfec227a495332595ce2b70c
Signed-off-by: Nico Huber <nico.huber@secunet.com>
Reviewed-on: http://review.coreboot.org/6250
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
Remove some ASCII art past 80 columns.
Change-Id: I00ad79f2e1ddd78935efcfab19d9e166f0349ae3
Signed-off-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-on: http://review.coreboot.org/6155
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
The objective here is to tighten coreboot up a bit by not repeating
common helpers. This makes the code base more consistent and
unified/tight.
Change-Id: Ia163eae68b4a84a00ed118125e70308fab1cea0c
Signed-off-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-on: http://review.coreboot.org/6215
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Implement the fix for the erratum #712. - Processor May Hang During Graphics Memory Controller
Sequencing
The processor may hang during a graphics memory controller (GMC) sleep state transitioning. The failure may
be processor specific and may be sensitive to temperature.
Potential Effect on System:
System hang.
Suggested Workaround:
BIOS should set D18F2x408_dct[1:0] bit 31 = 1b.
See Publication # 48931 Revision: 3.08
Change-Id: I4346fd4ef3cf554ffdaaad5ab6fc84e73532e885
Signed-off-by: Rudolf Marek <r.marek@assembler.cz>
Reviewed-on: http://review.coreboot.org/6216
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <gaumless@gmail.com>
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
In an abuild run, cbfstool is built in a shared directory
using "make tools". Unfortunately the build system doesn't
actually use that binary directly but creates a per-board
copy (for convenience purposes when editing the image later)
and uses that.
With this change the build system uses the original file but
still creates the copy for the user, avoiding the race while
ensuring convenience.
Change-Id: I38c603a7eca5ef859875ad3031bf7a850189645f
Signed-off-by: Patrick Georgi <patrick.georgi@secunet.com>
Reviewed-on: http://review.coreboot.org/6242
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Tested-by: build bot (Jenkins)
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Setting of `controller->reg_base` is of no use here, as it is never read
(in another function) later. Looks like this pattern originated from uhci.c
where it makes sense.
By removing the indirection through `reg_base` we also fix a possible
truncation to u32.
Change-Id: I5c99c5bf1f5b1d6c04bd84d87fd3e275fd7d0411
Signed-off-by: Nico Huber <nico.huber@secunet.com>
Reviewed-on: http://review.coreboot.org/6251
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Fix a possible null-pointer dereference (hopefully) before anyone runs
into this. Also don't switch ports to xHCI if initialization failed.
Change-Id: I5dbaeb435a98ead0b50d27fde13c9f1433ea3e81
Signed-off-by: Nico Huber <nico.huber@secunet.com>
Reviewed-on: http://review.coreboot.org/6245
Tested-by: build bot (Jenkins)
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
As the controller structure is never fully cleared, this one wasn't
initialized for non-pci controllers (but checked for non-null later).
Change-Id: I852671c5f55650bdb6cd97f4ec74b1f95ee894c7
Signed-off-by: Nico Huber <nico.huber@secunet.com>
Reviewed-on: http://review.coreboot.org/6246
Tested-by: build bot (Jenkins)
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Using void* for physical addresses leads to much casting and confuses
developers when to convert from physical to virtual addresses or
the other way around. When using plain integers for physical addresses
and pointers for virtual addresses things become much cleaner and we
won't ever end up dereferencing a physical address.
Change-Id: I24cd53b81c7863b6d14f0cbb4ce8937728b37c1c
Signed-off-by: Nico Huber <nico.huber@secunet.com>
Reviewed-on: http://review.coreboot.org/6244
Tested-by: build bot (Jenkins)
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Like done in FILO, libpayload's console drivers might be initialized
before a relocation. So keep physical pointers in there which won't
break on relocation.
Change-Id: I52e5d9d26801a53fd6a5f3c7ee03f61d6941d736
Signed-off-by: Nico Huber <nico.huber@secunet.com>
Reviewed-on: http://review.coreboot.org/6247
Tested-by: build bot (Jenkins)
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Remove a redundant phys_to_virt() that sneaked in the initialization of
PCI xHCI controllers. The use of casts from void* to u32 (and vice versa)
prompts for things going wrong here. That will be addressed in a later
commit.
Change-Id: Ibc71ed6ee7016529c0e3a51559aaec07aaaba315
Signed-off-by: Nico Huber <nico.huber@secunet.com>
Reviewed-on: http://review.coreboot.org/6243
Tested-by: build bot (Jenkins)
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Add the following useful macros:
* Absolute Value Macro
* Taking ceiling of (a / b)
* Check if value x is a power of 2 or not
Change-Id: I4e9a326aea3cdd963f13548d1fb63331a57d84b1
Signed-off-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-on: http://review.coreboot.org/6198
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
If one commented out HAVE_ACPI_RESUME in Kconfig file for a board
using agesa/hudson the build failed.
Change-Id: Ifbad8f6e23ce4b5431e596bf67e6ab108fedb4ce
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: http://review.coreboot.org/6253
Reviewed-by: Dave Frodin <dave.frodin@se-eng.com>
Tested-by: build bot (Jenkins)
Family15tn video bioses internal have a PCI ID of 1002/9901.
The vendor/device mapping in the family15tn/northbridge.c
file needs to map to 1002/9901 and not to 1002/9900.
This was tested on the amd/parmer mainboard.
Change-Id: I0153e9b522e847099c6054d91bf73b50966ed838
Signed-off-by: Dave Frodin <dave.frodin@se-eng.com>
Reviewed-on: http://review.coreboot.org/6252
Tested-by: build bot (Jenkins)
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
patch based on VMX support in intel/fsp_model_206ax and intel/model_6fx
tested/verified working on google/panther
Change-Id: I61232fdc2a29c53aa3bea5ea78b2fdc41fd7396a
Signed-off-by: Matt DeVillier <matt.devillier@gmail.com>
Reviewed-on: http://review.coreboot.org/6223
Tested-by: build bot (Jenkins)
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
ifdfake is the newest tool addition that leads to build time
races on highly parallel builds.
Change-Id: I86289e50079da851dcc8e1c05c2536d5c03de87c
Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Reviewed-on: http://review.coreboot.org/6197
Tested-by: build bot (Jenkins)
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
We have the macro, let us be sure to make use of it.
Change-Id: I8dc5ca580c7485e3cce7ebc29189a452de52b1b1
Signed-off-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-on: http://review.coreboot.org/6193
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
This code is not specific to any board or AGESA family.
Change-Id: I26c32fbe8e45018e239762b072dfe3da05271697
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: http://review.coreboot.org/5690
Tested-by: build bot (Jenkins)
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Whenever spi_xfer is called and whenver it's implemented, the natural unit for
the amount of data being transfered is bytes. The API expected things to be
expressed in bits, however, which led to a lot of multiplying and dividing by
eight, and checkes to make sure things were multiples of eight. All of that
can now be removed.
BUG=None
TEST=Built and booted on link, falco, peach_pit and nyan and looked for SPI
errors in the firmware log. Built for rambi.
BRANCH=None
Change-Id: I02365bdb6960a35def7be7a0cd1aa0a2cc09392f
Signed-off-by: Gabe Black <gabeblack@google.com>
Reviewed-on: https://chromium-review.googlesource.com/192049
Reviewed-by: Gabe Black <gabeblack@chromium.org>
Tested-by: Gabe Black <gabeblack@chromium.org>
Commit-Queue: Gabe Black <gabeblack@chromium.org>
[km: cherry-pick from chromium]
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: http://review.coreboot.org/6175
Tested-by: build bot (Jenkins)
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
The spi_flash_probe and and spi_setup_slave functions each took a max_hz
parameter and a spi_mode parameter which were never used.
BUG=None
TEST=Built for link, falco, rambi, nyan.
BRANCH=None
Change-Id: I3a2e0a9ab530bcc0f722f81f00e8c7bd1f6d2a22
Signed-off-by: Gabe Black <gabeblack@google.com>
Reviewed-on: https://chromium-review.googlesource.com/192046
Reviewed-by: Gabe Black <gabeblack@chromium.org>
Tested-by: Gabe Black <gabeblack@chromium.org>
Commit-Queue: Gabe Black <gabeblack@chromium.org>
[km: cherry-pick from chromium]
Signed-off-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Reviewed-on: http://review.coreboot.org/6174
Tested-by: build bot (Jenkins)
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
Duncan Laurie