If a section is bigger than the FD file it is injected into, and the FD
lies about the size of the FD file, ifdtool would crash because reading
in the section writes beyound the FD file in memory.
Change-Id: Idcfac2b1e2b5907fad34799e44a8abfd89190fcc
Signed-off-by: Stefan Reinauer <reinauer@google.com>
Reviewed-on: http://review.coreboot.org/1754
Tested-by: build bot (Jenkins)
Reviewed-by: Anton Kochkov <anton.kochkov@gmail.com>
This makes resume from S3 work again. The check is new and fails on
other boards, too.
Change-Id: I0ada569e4ba649b9ac82768b0888e16104c621e8
Signed-off-by: Nico Huber <nico.huber@secunet.com>
Reviewed-on: http://review.coreboot.org/1809
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
Detailed timing descriptor (DTD) is an 18 byte array describing video
mode (screen resolution, display properties, etc.) in Intel Option
ROM. Option ROM can support multiple video modes, specific mode is
picked by the BIOS through the appropriate Option ROM callback
function.
The new utility allows to interpret the 18 byte hex DTD dump, and/or
modify certain values, and generate a new DTD.
To parse the DTD contents just pass the 18 bytes to the utility in the
command line. To modify the existing contents and generate a new dump
precede the 18 bytes with '-m' and follow prompts.
Change-Id: Ib00bdaf42c350b98b5a48d08e6bb347b5ec25a8b
Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Reviewed-on: http://review.coreboot.org/1711
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
The endianness of an architecture is now set up automatically using Kconfig
and some common code. The available conversion functions were also expanded
to go to or from a particular endianness. Those use the abbreviation le or be
for little or big endian.
Built for Stumpy and saw coreinfo cbfs support work which uses network
byte order. Used the functions which convert to little endian to implement an
AHCI driver. The source arch is also little endian, so they were effectively
(and successfully) inert.
Change-Id: I3a2d2403855b3e0e93fa34f45e8e542b3e5afeac
Signed-off-by: Gabe Black <gabeblack@google.com>
Reviewed-on: http://review.coreboot.org/1719
Tested-by: build bot (Jenkins)
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
Give it somewhere to put the new info in sysinfo, and tell it how to parse
the new tables which it doesn't yet understand.
Change-Id: I01d3318138696e6407553c27c1814f79e3fbc4f8
Signed-off-by: Gabe Black <gabeblack@google.com>
Reviewed-on: http://review.coreboot.org/1718
Tested-by: build bot (Jenkins)
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
Read out the post code from the previous boot and
log it if the code is not one of the expected values.
Test:
1) interrupt the boot of the system, this is easiest
with warm reset button when servo is attached
2) check the event log with mosys
65 | 2012-09-09 12:32:11 | Last post code in previous boot | 0x9d
Change-Id: Id418f4c0cf005a3e97b8c63de67cb9a09bc57384
Signed-off-by: Duncan Laurie <dlaurie@chromium.org>
Reviewed-on: http://review.coreboot.org/1744
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
This will use 3 bytes of CMOS to keep track of the POST
code for the current boot while also leaving a record of
the previous boot.
The active bank is switched early in the bootblock.
Test:
1) clear cmos
2) reboot
3) use "mosys nvram dump" to verify that the first byte
contains 0x80 and the second byte contains 0xF8
4) powerd_suspend and then resume
5) use "mosys nvram dump" to verify that the first byte
contains 0x81 and the second byte contains 0xFD
Change-Id: I1ee6bb2dac053018f3042ab5a0b26c435dbfd151
Signed-off-by: Duncan Laurie <dlaurie@chromium.org>
Reviewed-on: http://review.coreboot.org/1743
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
This is a slight improvement over the rep movsb loop
Change-Id: Id71d9bfe5330b154a5c62fac85ce3955ae89b057
Signed-off-by: Stefan Reinauer <reinauer@google.com>
Reviewed-on: http://review.coreboot.org/1742
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
The linux kernel relies on the RTC reporting pending interrupts if
the RTC alarm was used to wake the system. If we clear these flags
here then the rtc-cmos driver in the kernel will think that no
interrupts are pending and will not re-start the timerqueue to
handle the alarm timerqueue node.
This flag doesn't exist in SMM but the rtc code is compiled there.
Since rtc_init() is not called by SMM it is guarded with an ifdef.
I performed several thousand suspend/resume cycles without seeing
an issue where hwclock was unable to read from /dev/rtc. There
still is a potential kernel issue where the timerqueue can stall
but this makes that much less likely to happen on resume.
Change-Id: I5a343da4ce5c4c8ec4783b4e503869ccfa5077f0
Signed-off-by: Duncan Laurie <dlaurie@chromium.org>
Reviewed-on: http://review.coreboot.org/1741
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
In case tseg_relocate() is called again on a pointer we should not
relocate it again.
Change-Id: Ida1f9c20dc94b448c773b14d8864afe585369119
Signed-off-by: Duncan Laurie <dlaurie@chromium.org>
Reviewed-on: http://review.coreboot.org/1740
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
We are seeing ME disabled and ME error events on some devices
and this extended info can help with debug.
Also fix a potential issue where if the log does manage to get
completely full it will never try to shrink it because the only
call to shrink the log happens after a successful event write.
Add a check at elog init time to shrink the log size.
Change-Id: Ib81dc231f6a004b341900374e6c07962cc292031
Signed-off-by: Duncan Laurie <dlaurie@chromium.org>
Reviewed-on: http://review.coreboot.org/1739
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
The handling of write enable was not entirely correct,
the opcode needs to be skipped when the controller is
locked down.
Addresses were not getting set properly for erase commands
which seemed to mostly work when the previous command had
set an address.
Tested by adding events to the event log at runtime on a
freslhy flashed device (with locked down SPI controller)
until the log log shrink happens to ensure it does not hang:
hexdump -C elog.event.kernel_clean
00000000 01 00 00 00 ad de 00 00 00 00
for x in $(seq 1 232); do
cat elog.event.kernel_clean > /sys/firmware/gsmi/append_to_eventlog
done
mosys eventlog list | tail -6
154 | 2012-09-01 13:54:43 | Kernel Event | Clean Shutdown
155 | 2012-09-01 13:54:43 | Kernel Event | Clean Shutdown
156 | 2012-09-01 13:54:43 | Kernel Event | Clean Shutdown
157 | 2012-09-01 13:54:43 | Kernel Event | Clean Shutdown
158 | 2012-09-01 13:54:43 | Log area cleared | 1030
159 | 2012-09-01 13:54:43 | Kernel Event | Clean Shutdown
Change-Id: I3a50dae54422a9ff37daefce3632f8bcbe4eb89f
Signed-off-by: Duncan Laurie <dlaurie@chromium.org>
Reviewed-on: http://review.coreboot.org/1717
Tested-by: build bot (Jenkins)
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
This is a cosmetic change which formats timestamp information
retrieved by cbmem.py.
Instead of printing timestamps in a single line, print them one per
line and add time (in us) elapsed since the previous timestamp.
time base 4149594, total entries 18
1:56,928
2:58,851 (1,923)
3:175,230 (116,378)
4:175,340 (109)
8:177,199 (1,859)
9:214,368 (37,168)
10:214,450 (81)
30:214,462 (11)
40:215,205 (743)
50:217,180 (1,974)
60:217,312 (132)
70:436,984 (219,671)
75:436,993 (8)
80:441,424 (4,431)
90:442,487 (1,062)
99:553,777 (111,289)
1000:556,513 (2,736)
1100:824,621 (268,107)
Change-Id: I0d25cafe766c10377017697e6b206276e1a92992
Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Reviewed-on: http://review.coreboot.org/1716
Tested-by: build bot (Jenkins)
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
For some reason which I fail to understand, specifying endiannes using
'@' (which means 'native' and should be the same as '<' on x86
platforms) causes cbmem.py to crash the machine on 64 bit systems.
What happens is that the addresses read from various table headers'
struct representations do not make sense, when bogus address gets
passed to get_phys_mem, the crash happens while that function is
executed.
dlaurie@ found out that replacing "@" with "<" in fact fixes the
issue. After some investigation I am just submitting this fix without
much understanding of the root cause.
Change-Id: Iaba9bc72a3f6b1d0407a5f1e3b459ccf5063969d
Signed-off-by: Duncan Laurie <dlaurie@chromium.org>
Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Reviewed-on: http://review.coreboot.org/1715
Tested-by: build bot (Jenkins)
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
When bringing up VGA by running the option rom it's sometimes
useful to get more information about the mode that gets set,
or the reason why the mode could not be set or a picture could
not be displayed. Also prefix the output from VBE mode setting
with VBE:
Copying VGA ROM Image from fff0fd78 to 0xc0000, 0x10000 bytes
Real mode stub @00000600: 867 bytes
Calling Option ROM...
int15_handler: INT15 function 5fac!
... Option ROM returned.
VBE: Getting information about VESA mode 4161
VBE: resolution: 1280x1024@16
VBE: framebuffer: d0000000
VBE: Setting VESA mode 4161
VGA Option ROM has been loaded
Change-Id: I2be11f095dc62ed3c99e0d4272ad9d6521608a44
Signed-off-by: Stefan Reinauer <reinauer@google.com>
Reviewed-on: http://review.coreboot.org/1714
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
CBFS allows coreboot rom images that are only partially covered
by the filesystem itself. The intention of this feature was to
allow EC / ME / IMC firmware to be inserted easily at the beginning
of the image. However, this was never implemented in cbfstool.
This patch implements an additional parameter for cbfstool.
If you call cbfstool like this:
cbfstool coreboot.rom create 8192K bootblock.bin 64 0x700000
it will now create an 8M image with CBFS covering the last 1M of
that image.
Test:
cbfstool coreboot.rom create 8192K bootblock.bin 64 0x700000
creates an 8M image that is 7M of 0xff and 1M of CBFS.
Change-Id: I5c016b4bf32433f160b43f4df2dd768276f4c70b
Signed-off-by: Stefan Reinauer <reinauer@google.com>
Reviewed-on: http://review.coreboot.org/1708
Tested-by: build bot (Jenkins)
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
Now that we have FMAP support in coreboot use it to find the
offset in flash for ELOG to use.
If coreboot has elog configured with a smaller size then use
that over the FMAP size. This is because I set aside a 16KB
region in the FMAP but we only use 4KB of it to keep the impact
to boot/resume speed to a minimum.
FMAP: Found "FMAP" version 1.0 at ffe10000.
FMAP: base = 0 size = 800000 #areas = 32
FMAP: area RW_ELOG found
FMAP: offset: 3f0000
FMAP: size: 16384 bytes
FMAP: No valid base address, using 0xff800000
ELOG: base=0x003f0000 base_ptr=0xffbf0000
ELOG: MEM @0x00190ad8 FLASH @0xffbf0000
ELOG: areas are 4096 bytes, full threshold 3072, shrink size 1024
Change-Id: I3d826812c0f259d61f41b42797c58dd179f9f1c8
Signed-off-by: Duncan Laurie <dlaurie@chromium.org>
Reviewed-on: http://review.coreboot.org/1706
Tested-by: build bot (Jenkins)
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
Now that WREN prefix is handled properly ELOG is able to write
when the SPI controller is locked down.
To test, ensure that runtime SPI write via ELOG is successful by
checking the event log for a kernel shutdown reason code:
5 | 2012-08-27 11:09:48 | Kernel Event | Clean Shutdown
6 | 2012-08-27 11:09:50 | System boot | 26
7 | 2012-08-27 11:09:50 | System Reset
Change-Id: If6d0dced7cb0f5ca7038b3d758f31b856826d30b
Signed-off-by: Duncan Laurie <dlaurie@chromium.org>
Reviewed-on: http://review.coreboot.org/1712
Reviewed-by: Marc Jones <marcj303@gmail.com>
Tested-by: build bot (Jenkins)
The code that attempts to use the opmenu needs to have a special
case for write enable now that it is handled as an atomic prefix
and not as a standalone opcode.
To test, ensure that runtime SPI write via ELOG is successful by
checking the event log for a kernel shutdown reason code:
5 | 2012-08-27 11:09:48 | Kernel Event | Clean Shutdown
6 | 2012-08-27 11:09:50 | System boot | 26
7 | 2012-08-27 11:09:50 | System Reset
Change-Id: I527638ef3e2a5ab100192c5be6e6b3b40916295a
Signed-off-by: Duncan Laurie <dlaurie@chromium.org>
Reviewed-on: http://review.coreboot.org/1710
Reviewed-by: Marc Jones <marcj303@gmail.com>
Tested-by: build bot (Jenkins)
Check the RTC on boot after RTC battery failure and ensure
that the reported build date matches what is reported:
> grep ^rtc /proc/driver/rtc
rtc_time : 01:00:21
rtc_date : 2012-08-16
Change-Id: If23f436796754c68ae6244ef7633ff4fa0a93603
Signed-off-by: Duncan Laurie <dlaurie@chromium.org>
Reviewed-on: http://review.coreboot.org/1709
Reviewed-by: Marc Jones <marcj303@gmail.com>
Tested-by: build bot (Jenkins)
The "debug" macro used internally in the libpayload USB subsystem was very
generically named and would leak into consumers of the library that included
usb.h directly or indirectly. This change turns that #define from a macro into
a static inline function to move away from the preprocessor, and also renames
it to usb_debug so it's less likely to collide with something unrelated.
Change-Id: I18717df111aa9671495f8a2a5bdb2c6311fa7acf
Signed-off-by: Gabe Black <gabeblack@google.com>
Reviewed-on: http://review.coreboot.org/1738
Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
Tested-by: build bot (Jenkins)
While it might be slightly more convenient to not have to call usb_poll
manually after calling usb_initialize, you'll still likely want to call it
before trying to use a USB device since one have have been hotplugged since
you last looked. By not calling usb_poll, usb_initialize completes quickly and
can be called unconditionally without a long delay. The delay can be put off
until later when we're sure it's necessary.
Change-Id: Ib8b1bdea996702c42d1b7021f492d9f8e174d304
Signed-off-by: Gabe Black <gabeblack@google.com>
Reviewed-on: http://review.coreboot.org/1737
Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
Tested-by: build bot (Jenkins)
The usb_initialize function would scan for USB host controllers by brute force
iterating over all possible busses, devices, and functions. This change makes
it recursively scan busses only if it finds them on the other side of a bridge,
and only scan for functions beyond function 0 if the device claims to be
multifunction.
This change also takes the opportunity to clean up some style problems
throughout the file.
Change-Id: I0f5e8b9a454a42a76d30bccca898c8e1af770b2b
Signed-off-by: Gabe Black <gabeblack@google.com>
Reviewed-on: http://review.coreboot.org/1736
Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
Tested-by: build bot (Jenkins)
Change-Id: I02cf353ce7c955cb11ca11c0d5b8aa630cf15fdb
Signed-off-by: Gabe Black <gabeblack@google.com>
Reviewed-on: http://review.coreboot.org/1735
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
gcc recognizes the format function attribute which tells the compiler to expect
the format string to look a certain way and for its arguments to be of
appropriate types. This helps to prevent errors like the one that was recently
fixed in libpayload's assert.
Change-Id: I284ae8bff32f72cfd2d1a250d126c729b38a5730
Signed-off-by: Gabe Black <gabeblack@google.com>
Reviewed-on: http://review.coreboot.org/1731
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
The assert macro in libpayload was using a format string which printed the
line number with %s. The line number came from the __LINE__ predefined macro
which resolves to an integer constant.
Change-Id: I0e00d42a1569802137cf440af3061d7f397fdd27
Signed-off-by: Gabe Black <gabeblack@google.com>
Reviewed-on: http://review.coreboot.org/1730
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
Change-Id: I0f3a82de860fd3afa10a557b37fb90fe6b06ae90
Signed-off-by: Gabe Black <gabeblack@google.com>
Reviewed-on: http://review.coreboot.org/1726
Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
Tested-by: build bot (Jenkins)
The implementations for various stdlib functions in libc/memory.c are very
generic and should work under just about any circumstances. They are
unfortunately also very slow. This change makes them weak symbols so that
faster versions can be defined on a per architecture basis which will
automatically take the place of the slow versions.
Change-Id: Ia1ac90d9dcd45962b2a15f61ecc74b0a4676048d
Signed-off-by: Gabe Black <gabeblack@google.com>
Reviewed-on: http://review.coreboot.org/1725
Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
Tested-by: build bot (Jenkins)
This information is now stored in a structure instead of in a few seperate
fields. libpayload hadn't been updated to reflect the new layout or to consume
the new information intelligently.
Change-Id: Ice3486ffcdcdbe1f16f9c84515120c591d8dc882
Signed-off-by: Gabe Black <gabeblack@google.com>
Reviewed-on: http://review.coreboot.org/1724
Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
Tested-by: build bot (Jenkins)
Change-Id: Ifb7c18f9ca566bd50ca138ffd8af951375089537
Signed-off-by: Gabe Black <gabeblack@google.com>
Reviewed-on: http://review.coreboot.org/1722
Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
Tested-by: build bot (Jenkins)
These end up being loaded at 0 otherwise and overwrite some coreboot tables.
Built and booted on Stumpy. Saw that the coreboot tables were no longer
overwritten.
Change-Id: Ia9f521d976d0ad544a8205323ae0ddfa8d253d29
Signed-off-by: Gabe Black <gabeblack@google.com>
Reviewed-on: http://review.coreboot.org/1721
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
This file uses uint*_t types but hadn't included stdint.h itself.
Change-Id: Ib883f62951bae1ece5134c6bd0f4799a80740e8e
Signed-off-by: Gabe Black <gabeblack@google.com>
Reviewed-on: http://review.coreboot.org/1720
Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
Tested-by: build bot (Jenkins)
cbfstool was not looking at any dependencies when building
by running make in util/cbfstool. By fixing this it's not
required to make clean every time you edit a file in there.
Change-Id: I544fd54d4b9dd3b277996c21ade56dc086b84800
Signed-off-by: Stefan Reinauer <reinauer@google.com>
Reviewed-on: http://review.coreboot.org/1707
Reviewed-by: Marc Jones <marcj303@gmail.com>
Tested-by: build bot (Jenkins)
This appears to fix an infrequent resume hang on Ivybridge.
Tested on 2 devices with 15k suspend/resume cycles each
Change-Id: I53618bc7966824413f1720a2be3cbd2550e29473
Signed-off-by: Duncan Laurie <dlaurie@chromium.org>
Reviewed-on: http://review.coreboot.org/1704
Reviewed-by: Marc Jones <marcj303@gmail.com>
Tested-by: build bot (Jenkins)
(elog portion, support in EC code pending)
- Use a new EC command to read the last post code
from the previous boot
- If the post code is not well-known final boot
or resume code then log it
Change-Id: Id6249e9a182243eb87c777edd56f48de72125e77
Signed-off-by: Duncan Laurie <dlaurie@chromium.org>
Reviewed-on: http://review.coreboot.org/1703
Reviewed-by: Marc Jones <marcj303@gmail.com>
Tested-by: build bot (Jenkins)
Until now, the MRC cache position and size was hard coded in Kconfig.
However, on ChromeOS devices, it should be determined by reading the
FMAP.
This patch provides a minimalistic FMAP parser (libflashmap was too
complex and OS centered) to allow reading the in-ROM flash map and
look for sections.
This will also be needed on some partner devices where coreboot will
have to find the VPD in order to set up the device's mac address
correctly.
The MRC cache implementation demonstrates how to use the FMAP parser.
Change-Id: I34964b72587443a6ca4f27407d778af8728565f8
Signed-off-by: Stefan Reinauer <reinauer@google.com>
Reviewed-on: http://review.coreboot.org/1701
Reviewed-by: Marc Jones <marcj303@gmail.com>
Tested-by: build bot (Jenkins)
Otherwise object paths will look like build/cbfs/"fallback"/...
Change-Id: I3e60f90f7490e71b0da075d3ea8fc847abc07938
Signed-off-by: Stefan Reinauer <reinauer@google.com>
Reviewed-on: http://review.coreboot.org/1700
Reviewed-by: Marc Jones <marcj303@gmail.com>
Tested-by: build bot (Jenkins)
This is purely cosmetic. All error messages in the Sandybridge raminit
code printed a newline at the end.
Change-Id: I880d291928291d487039850a2a3d53a1101124ba
Signed-off-by: Stefan Reinauer <reinauer@google.com>
Reviewed-on: http://review.coreboot.org/1699
Reviewed-by: Marc Jones <marcj303@gmail.com>
Tested-by: build bot (Jenkins)
build.h is generated at build time,
with highly parallel builds, we might try to compile the rtc driver too
early.
Change-Id: I9a2681484d58b67ed3061669fbdf52ac5ad14dab
Signed-off-by: Vincent Palatin <vpalatin@chromium.org>
Reviewed-on: http://review.coreboot.org/1698
Reviewed-by: Marc Jones <marcj303@gmail.com>
Tested-by: build bot (Jenkins)
When a power failure happens on the RTC rail, the CMOS memory (including
the RTC registers) is filled with garbage.
So, we erase the full first bank (112 bytes) and we reset the RTC date
to the build date.
To test, disconnect the CMOS battery to produce an RTC power
failure, then boot the machine and observe the RTC date is the build
date using "cat /sys/class/rtc/rtc0/date"
Change-Id: I684bb3ad5079f96825555d4ed84dc0f7914e9884
Signed-off-by: Vincent Palatin <vpalatin@chromium.org>
Reviewed-on: http://review.coreboot.org/1697
Reviewed-by: Marc Jones <marcj303@gmail.com>
Tested-by: build bot (Jenkins)
Each G34 socket has two node. Previous lapic algorithm is written for
the CPU which has one node per socket. I test the code on h8qgi with
4 family 15 CPUs(8 cores per CPU). The topology is:
socket 0 --> Node 0, Node 1
socket 2 --> Node 2, Node 3
socket 1 --> Node 4, Node 5
socket 3 --> Node 6, Node 7
Each node has 4 cores.
I change the code according to this topology.
Change-Id: I45f242e0dfc61bd9b18afc952d7a0ad6a0fc3855
Signed-off-by: Siyuan Wang <SiYuan.Wang@amd.com>
Signed-off-by: Siyuan Wang <wangsiyuanbuaa@gmail.com>
Reviewed-on: http://review.coreboot.org/1659
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marcj303@gmail.com>
This will allow the lower bank to be cleared without impacting the
ability to suspend/resume.
Change-Id: Iaec3c9e7e40c334053c814eaddd1f614df245a73
Signed-off-by: Duncan Laurie <dlaurie@chromium.org>
Reviewed-on: http://review.coreboot.org/1696
Reviewed-by: Marc Jones <marcj303@gmail.com>
Tested-by: build bot (Jenkins)
On Panther Point PCH (and maybe cougar point), when some of the register
D reserved bits are set, the RTC starts misbehaving (e.g. incrementing
the year byte every second).
There are probably undocumented features implemented behind those bits.
Let's reset register D to a known state to ensure we get the expected
RTC behavior.
Change-Id: I7e2c2a2c6130a974bccb3d760b41eaa579a58b67
Signed-off-by: Vincent Palatin <vpalatin@chromium.org>
Reviewed-on: http://review.coreboot.org/1695
Reviewed-by: Marc Jones <marcj303@gmail.com>
Tested-by: build bot (Jenkins)
We always define CONFIG_ variables, even if they're not set.
Hence, remove the check whether CONFIG_UDELAY_TIMER2 is defined
Change-Id: Iefdf2389941f2cc63ae4f13ac6b213da4c96b201
Signed-off-by: Stefan Reinauer <reinauer@google.com>
Reviewed-on: http://review.coreboot.org/1694
Reviewed-by: Marc Jones <marcj303@gmail.com>
Tested-by: build bot (Jenkins)
Right now we only had a post code for "All devices enabled" which
was emitted at the wrong time (after the device initialize stage
rather than the device enable stage)
Change-Id: Iee82bff020de844c7095703f8d6521953003032c
Signed-off-by: Stefan Reinauer <reinauer@google.com>
Reviewed-on: http://review.coreboot.org/1693
Reviewed-by: Marc Jones <marcj303@gmail.com>
Tested-by: build bot (Jenkins)
Our linker script for romstage checks for global variables and
makes the build fail if there are any (on non-AMD systems).
This is great, but having the build fail without any indication
which variables are global is not very useful.
Moving the check to the Makefile allows us to let the linking stage
succeed and reveil which variable names end up in the data and bss
sections of the binary.
To test, add "int foo;" as the first line in src/mainboard/samsung/lumpy/romstage.c
and build coreboot for Lumpy. See the build break the following
way:
LINK cbfs/fallback/romstage_null.debug
Forbidden global variables in romstage:
00006a84 B foo
Change-Id: I3c8780888f46a6577ffd36bcea317997b4f84f6f
Signed-off-by: Stefan Reinauer <reinauer@google.com>
Reviewed-on: http://review.coreboot.org/1692
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marcj303@gmail.com>
To allow easy experimentation with thermals, leave power control
registers unlocked.
Change-Id: Ia53065f3f220c2faed58e7d53e60c3f169ae58ec
Signed-off-by: Sameer Nanda <snanda@chromium.org>
Reviewed-on: http://review.coreboot.org/1688
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marcj303@gmail.com>
Every line of text after a 'help' label in a Kconfig
file must have the same whitespace preceding it, otherwise
it's no longer considered help text.
Change-Id: I97093bee72b295b315d78d4c26d7186bf1017fda
Signed-off-by: Dave Frodin <dave.frodin@se-eng.com>
Reviewed-on: http://review.coreboot.org/1687
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
Tested-by: build bot (Jenkins)
PCIE devices are detected and initialized by the AMD PCIe init functions,
which is in cimx rd890. The parameters are read from devicetree.cb before PCIe init.
Now, all bridges and devices are trained on the device 0.0 enable.
After PCIe init, the PCIe ports with devices are on and the PCIe ports
without devices are off. so resources may be allocated correctly
during the rest of the PCI scan.
But if the devicetree was being used to enable/disable devices after initialization,
the problems would arise. Take a look at the serial log:
do_pci_scan_bridge for PCI: 00:02.0
PCI: pci_scan_bus for bus 01
PCI: pci_scan_bus returning with max=001
do_pci_scan_bridge returns max 1
do_pci_scan_bridge for PCI: 00:03.0
PCI: pci_scan_bus for bus 02
PCI: pci_scan_bus returning with max=002
do_pci_scan_bridge returns max 2
do_pci_scan_bridge for PCI: 00:04.0
PCI: pci_scan_bus for bus 03
PCI: pci_scan_bus returning with max=003
do_pci_scan_bridge returns max 3
PCI bridge 02.0, 03.0 and 04.0 are not inserted devices, but these bridges
are still scanned. This is not correct.
Change-Id: I87dac5f062c6926081970ed0c5f26a7e3f447395
Signed-off-by: Siyuan Wang <SiYuan.Wang@amd.com>
Signed-off-by: Siyuan Wang <wangsiyuanbuaa@gmail.com>
Reviewed-on: http://review.coreboot.org/1640
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marcj303@gmail.com>
There are four mainboards using agesa family15 code:
Supermicro h8scm and h8qgi, Tyan s8226 and AMD dinar.
All of these boards' PCI domain starts from 0x18.0. Take h8scm as
an example, PCI devices from 0.0 to 0x14.5 is under 0x18.0.
Now, the PCI domain's scan bus function stats from 0.0. This would
result to the PCI devices be scanned twice. Because when the function
run to device 18.0, it would scan from 0.0 again.
This issue would result to 2 problems:
1) PCI device may be assigned two different PCI address.
If this happenned on VGA device, coreboot maybe not load
vga bios correctly.
2) coreboot initializes rd890's IO APIC twice.
So this patch scans from 0x18.0 and could resolve the problems above.
Change-Id: I90fbdf695413fd24c7a5e3e9b426dc7ca6e128b1
Signed-off-by: Siyuan Wang <SiYuan.Wang@amd.com>
Signed-off-by: Siyuan Wang <wangsiyuanbuaa@gmail.com>
Reviewed-on: http://review.coreboot.org/1639
Tested-by: build bot (Jenkins)
Reviewed-by: Marc Jones <marcj303@gmail.com>
Stefan Reinauer