Some LPC initialiation can save some lines of code when being able
to use the functions `io_apic_read()` and `io_apic_write()`.
As these two functions are now public, remove them from the generic
driver as otherwise we get a build errors like the following.
[…]
Building roda/rk9; i386: ok, using i386-elf-gcc
Using payload /srv/jenkins/payloads/seabios/bios.bin.elf
Creating config file... (blobs, ccache) ok; Compiling image on 4 cpus in parallel .. FAILED after 12s!
Log excerpt:
coreboot-builds/roda_rk9/arch/x86/lib/ramstage.o: In function `io_apic_write':
/srv/jenkins/.jenkins/jobs/coreboot-gerrit/workspace/src/arch/x86/lib/ioapic.c:32: multiple definition of `io_apic_write'
coreboot-builds/roda_rk9/drivers/generic/ioapic/ramstage.o:/srv/jenkins/.jenkins/jobs/coreboot-gerrit/workspace/src/drivers/generic/ioapic/ioapic.c:22: first defined here
collect2: error: ld returned 1 exit status
make: *** [coreboot-builds/roda_rk9/generated/coreboot_ram.o] Error 1
make: *** Waiting for unfinished jobs....
[…]
Change-Id: Id600007573ff011576967339cc66e6c883a2ed5a
Signed-off-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-on: http://review.coreboot.org/3180
Tested-by: build bot (Jenkins)
Reviewed-by: Aaron Durbin <adurbin@google.com>
Internally there were states that had an attribute to
indicate that the timers needed to be drained. Now that
there is a way to block state transitions rely on this
ability instead of draining timers. The timers will
drain themselves when a state is blocked.
Change-Id: I59be9a71b2fd5a17310854d2f91c2a8957aafc28
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/3205
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
In order to properly sequence the boot state machine it's
important that outside code can block the transition from
one state to the next. When timers are not involved there's
no reason for any of the existing code to block a state
transition. However, if there is a timer callback that needs to
complete by a certain point in the boot sequence it is necessary
to place a block for the given state.
To that end, 4 new functions are added to provide the API for
blocking a state.
1. boot_state_block(boot_state_t state, boot_state_sequence_t seq);
2. boot_state_unblock(boot_state_t state, boot_state_sequence_t seq);
3. boot_state_current_block(void);
4. boot_state_current_unblock(void);
Change-Id: Ieb37050ff652fd85a6b1e0e2f81a1a2807bab8e0
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/3204
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
When the haswell MP/SMM code was developed it was using a coreboot
repository that did not contain the asmlinkage macro. Now that the
asmlinkage macro exists use it.
BUG=None
BRANCH=None
TEST=Built and booted.
Change-Id: I662f1b16d1777263b96a427334fff8f98a407755
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/3203
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
We have the monotonic timer implemented on exynos now, and this
also enables helpful bootstage prints with timing info.
Change-Id: I3baa4c9d70d4b4d059abd5e05eddcabd5258dbfd
Signed-off-by: David Hendricks <dhendrix@chromium.org>
Reviewed-on: http://review.coreboot.org/3210
Reviewed-by: Aaron Durbin <adurbin@google.com>
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Some boards use the local apic for udelay(), but they also provide
their own implementation of udelay() for SMM. The reason for using
the local apic for udelay() in ramstage is to not have to pay the
penalty of calibrating the TSC frequency. Therefore provide a
TSC_CONSTANT_RATE option to indicate that TSC calibration is not
needed. Instead rely on the presence of a tsc_freq_mhz() function
provided by the cpu/board. Additionally, assume that if
TSC_CONSTANT_RATE is selected the udelay() function in SMM will
be the tsc.
Change-Id: I1629c2fbe3431772b4e80495160584fb6f599e9e
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/3168
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
Instead of using the local apic timer for udelay() use the tsc.
That way SMM, romstage, and ramstage all use the same delay
functionality.
Change-Id: I024de5af01eb5de09318e13d0428ee98c132f594
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/3169
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
The stack used on the ASRock E350M1 is significantly less than
what we currently set (64k per core). In fact, we use about half
of the default stack size (4k) on core 0 and even less on non
BSP cores [1]:
$ grep stack coreboot_without_patch_but_monotonic_timer.log
CPU1: stack_base 002a0000, stack_end 002afff8
CPU1: stack: 002a0000 - 002b0000, lowest used address 002afda8, stack used: 600 bytes
CPU0: stack: 002b0000 - 002c0000, lowest used address 002bf75c, stack used: 2212 bytes
Removing the Kconfig variable STACK_SIZE to use the default results
in the following numbers of stack usage.
$ grep stack coreboot_with_patch.log
CPU1: stack_base 00287000, stack_end 00287ff8
CPU1: stack: 00287000 - 00288000, lowest used address 00287da8, stack used: 600 bytes
CPU0: stack: 00288000 - 00289000, lowest used address 0028875c, stack used: 2212 bytes
[1] http://review.coreboot.org/#/c/3154/
(comment May 2 10:21 AM)
Change-Id: Ibdb2102c86094fce3787e3b5a162ca8423de205c
Signed-off-by: Stefan Reinauer <reinauer@google.com>
Tested-by: Paul Menzel <paulepanter@users.sourceforge.net>
Signed-off-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-on: http://review.coreboot.org/3209
Tested-by: build bot (Jenkins)
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
The cbfs core code would print out all unmatched file
names when searching for a file. This contributes to a lot
of unnecessary messages in the boot log. Change this
message to a DEBUG one so that it will only be printed when
CONFIG_DEBUG_CBFS is enabled.
Change-Id: I34c747e0d3406351318abf70994dbc0bb3fa6c01
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/3164
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Marc Jones <marc.jones@se-eng.com>
Tested-by: build bot (Jenkins)
This re-introduces 2fde966 (http://review.coreboot.org/#/c/3177/)
which was reverted due to unsatisfied dependencies.
time.h We Hardly Knew Ye.
This deprecates time.h which is currently only used by Exynos5250 and
Snow. The original idea was to try and unify some of the various timer
interfaces and has been supplanted by the monotonic timer API.
timer_us() is now obsolete. timer_start() is now mct_start() and
is exposed in exynos5250/clk.h.
Change-Id: I8e60105629d9da68ed622e89209b3ef6c8e2445b
Signed-off-by: David Hendricks <dhendrix@chromium.org>
Reviewed-on: http://review.coreboot.org/3201
Tested-by: build bot (Jenkins)
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
The current way to get a simple mono_time difference is:
1. Declare a rela_time struct
2. Assign it the value of mono_time_diff(t1, t2)
3. Get microseconds from it using rela_time_in_microseconds().
This patch adds a simpler method. Now one only needs to call
mono_time_diff_microseconds(t1, t2) to obtain the same value which
is produced from the above three steps.
Change-Id: Ibfc9cd211e48e8e60a0a7703bff09cee3250e88b
Signed-off-by: David Hendricks <dhendrix@chromium.org>
Reviewed-on: http://review.coreboot.org/3190
Tested-by: build bot (Jenkins)
This goes thru various call sites where we used timer_us() and updates
them to use the new monotonic timer API.
udelay() changed substantially and now gracefully handles wraparound.
Change-Id: Ie2cc86a4125cf0de12837fd7d337a11aed25715c
Signed-off-by: David Hendricks <dhendrix@chromium.org>
Reviewed-on: http://review.coreboot.org/3176
Reviewed-by: Aaron Durbin <adurbin@google.com>
Tested-by: build bot (Jenkins)
1. Move comment for console init to correct place.
2. Start output with capital letter and add full stop at the end.
3. Add missing »)« at the end of description of GPIO 10.
4. Use tabulators instead of spaces.
5. Indent the code automatically using GNU indent [1] with the `-sc`
switch adding stars in front of comment blocks as the good indent
manual documents.
$ indent -linux -sc src/mainboard/lenovo/x60/romstage.c
Leave the numbers left aligned as it is more beneficial to be
able to run indent without adapting the result afterward.
[1] http://www.coreboot.org/Development_Guidelines#Coding_Style
Change-Id: I2fa018ec28ff19d23d68754b565c13a7d7a57355
Signed-off-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-on: http://review.coreboot.org/3185
Tested-by: build bot (Jenkins)
Reviewed-by: Denis Carikli <GNUtoo@no-log.org>
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
This reverts commit 2fde9668b4
Somehow this got merged before its dependencies. 3190 must be merged first, followed by 3176. However 3190 will fail while this patch is in. So the situation can't correct itself.
Reverting this until the other two go in.
Change-Id: I176f37c12711849c96f1889eacad38c00a8142c4
Signed-off-by: David Hendricks <dhendrix@chromium.org>
Reviewed-on: http://review.coreboot.org/3195
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
If the SD controller is "off" hudson.c won't disable that because,
there is no code for this yet.
The PCI device is still visible and PCI BAR will be allocated
by Linux. Unfortunately it may happen that the particular address
is used by non-standard BAR for SPI controller.
Change-Id: Ied7c581727541e2c81b0b1c2b70fd32de0014730
Signed-off-by: Rudolf Marek <r.marek@assembler.cz>
Reviewed-on: http://review.coreboot.org/3167
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
Fix Warning:
cpuFeatureLeveling.c:265, GNU Compiler 4 (gcc), Priority: Normal
cast to pointer from integer of different size [-Wint-to-pointer-cast]
with an intermediate cast to (intptr_t)
Change-Id: I3bfd2ea1e797632316675338789dabef8f73ba64
Signed-off-by: Martin Roth <martin.roth@se-eng.com>
Reviewed-on: http://review.coreboot.org/3126
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Dave Frodin <dave.frodin@se-eng.com>
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
This fixes 3 warnings in the Proc/Common directory:
AmdS3Save.c:250, GNU Compiler 4 (gcc), Priority: Normal
AmdS3LateRestore.c:123, GNU Compiler 4 (gcc), Priority: Normal
cast from pointer to integer of different size [-Wpointer-to-int-cast]
Fixed with a second cast to (intptr_t)
AmdInitReset.c:153, GNU Compiler 4 (gcc), Priority: Normal
statement with no effect [-Wunused-value]
Fixed by commenting the line out as it is in the other families code.
Change-Id: Ib35ec466671712af01568b7c2a18ee138fe883c0
Signed-off-by: Martin Roth <martin.roth@se-eng.com>
Reviewed-on: http://review.coreboot.org/3125
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Dave Frodin <dave.frodin@se-eng.com>
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
We write CMOS data to 128 byte files, which is a problem
when using them later-on (eg. as part of a coreboot image)
where nvramtool assumes them to be 256 byte, and so data
corruption occurs.
Change-Id: Ibc919c95f6d522866b21fd313ceb023e73d09fb9
Signed-off-by: Patrick Georgi <patrick.georgi@secunet.com>
Reviewed-on: http://review.coreboot.org/3186
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
time.h We Hardly Knew Ye.
This deprecates time.h which is currently only used by Exynos5250 and
Snow. The original idea was to try and unify some of the various timer
interfaces and has been supplanted by the monotonic timer API.
timer_us() is now obsolete. timer_start() is now mct_start() and
is exposed in exynos5250/clk.h.
Signed-off-by: David Hendricks <dhendrix@chromium.org>
Change-Id: I14ebf75649d101491252c9aafea12f73ccf446b5
Reviewed-on: http://review.coreboot.org/3177
Reviewed-by: Aaron Durbin <adurbin@google.com>
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
Update crossgcc to use gcc 4.7.3
The resulting coreboot.rom is not runtime tested (any volunteers?).
Drop the texinfo patch, rename the armv7a patch.
Some Linux distributions have moved on to gcc 4.8,
under certain circumstances this version can't (cross-)compile gcc 4.7.2
Bug report: http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56927
Change-Id: Id8ce5f86c34e1a0900d44dc6ae4e81cb9548ecc2
Signed-off-by: Idwer Vollering <vidwer@gmail.com>
Reviewed-on: http://review.coreboot.org/3112
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Martin Roth <martin.roth@se-eng.com>
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
Commit
commit 825c78b5da
Author: David Hubbard <david.c.hubbard+coreboot@gmail.com>
Date: Thu May 2 18:06:03 2013 -0600
mainboard/{asus/f2a85-m,amd/thatcher}: move UDELAY_LAPIC
Reviewed-on: http://review.coreboot.org/3178
adds `UDELAY_LAPIC` to `cpu/amd/agesa/family15tn/Kconfig`. This is
not needed, because since commit
commit e135ac5a7e
Author: Patrick Georgi <patrick.georgi@secunet.com>
Date: Tue Nov 20 11:53:47 2012 +0100
Remove AMD special case for LAPIC based udelay()
Reviewed-on: http://review.coreboot.org/1618
`select UDELAY_LAPIC` is present in `src/cpu/amd/agesa/Kconfig` which
applies also to AMD Family 15tn.
Therefore remove `select UDELAY_LAPIC` again from
`cpu/amd/agesa/family15tn/Kconfig`.
Change-Id: I98b783a97c4a1e45ecb29b776cb3d3877bad9c0f
Signed-off-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-on: http://review.coreboot.org/3179
Tested-by: build bot (Jenkins)
Reviewed-by: Peter Stuge <peter@stuge.se>
This implements the new monotonic timer API using the global
multi-core timer (MCT).
Change-Id: Id56249ff5d3e0f85808f5754954c83c0bc75f1c1
Signed-off-by: David Hendricks <dhendrix@chromium.org>
Reviewed-on: http://review.coreboot.org/3175
Reviewed-by: Aaron Durbin <adurbin@google.com>
Tested-by: build bot (Jenkins)
The following command was used to correct all occurences of this typo.
$ git grep -l "them implem" | xargs sed -i 's/them implem/then implem/'
Change-Id: Iebd4635867d67861aaf4d4d64ca8a67e87833f38
Signed-off-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-on: http://review.coreboot.org/3145
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
Commit »haswell: Add initial support for Haswell platforms« (76c3700f)
[1] used `1 << 25` to set the I/O APIC ID of 2. Instead using
`2 << 24`, which is the same value, makes it clear, that the
I/O APIC ID is 2.
Commit »Intel Panther Point PCH: Use 2 << 24 to clarify that APIC ID
is 2« (8c937c7e) [2] is used as a template.
[1] http://review.coreboot.org/2616
[2] http://review.coreboot.org/3100
Change-Id: I28f9e90856157b4fdd9a1e781472cc4f51d25ece
Signed-off-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-on: http://review.coreboot.org/3123
Tested-by: build bot (Jenkins)
Reviewed-by: Aaron Durbin <adurbin@google.com>
Capitalizing CBMEM seems to be the official spelling as can be seen
in the descriptions around the `EARLY_CBMEM_INIT` Kconfig option.
Change-Id: I046a678c3b04ef7e681de46aa137cedc405d546f
Signed-off-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-on: http://review.coreboot.org/3143
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
The cbfs core code would print out all unmatched file
names when searching for a file. This contributes to a lot
of unnecessary messages in the boot log. Change this
message to a DEBUG one so that it will only be printed when
CONFIG_DEBUG_CBFS is enabled.
Change-Id: I1e46a4b21d80e5d2f9b511a163def7f5d4e0fb99
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/3131
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Marc Jones <marc.jones@se-eng.com>
Tested-by: build bot (Jenkins)
Stefan Reinauer suggested 'select UDELAY_LAPIC' did not belong in
f2a85-m/Kconfig. It got there via copy-paste from thatcher/Kconfig
so this commit removes the 'select UDELAY_LAPIC' from both and puts
it in cpu/amd/agesa/family15tn/Kconfig
Since f2a85-m is the only Thatcher board coreboot supports right
now, this should not break any other boards.
Change-Id: I811b579c31f8d259a237d3a6724ad3b17f3a6c3e
Signed-off-by: David Hubbard <david.c.hubbard+coreboot@gmail.com>
Reviewed-on: http://review.coreboot.org/3178
Reviewed-by: Peter Stuge <peter@stuge.se>
Tested-by: build bot (Jenkins)
The old approach was to invalidate the entire TLB every time we set up
a table entry. This worked because we didn't turn the MMU on until
after we had set everything up. This patch uses the TLBIMVAA wrapper
to invalidate each entry as it's added/modified.
Change-Id: I27654a543a2015574d910e15d48b3d3845fdb6d1
Signed-off-by: David Hendricks <dhendrix@chromium.org>
Reviewed-on: http://review.coreboot.org/3166
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
Tested-by: build bot (Jenkins)
The "gigabit ethernet controller" (GEC) block was added to AMD
Hudson A55E to integrate ethernet capabilities into an AMD
southbridge.
The GEC is designed to work with B50610 and B50610M gigabit PHY
chips from Broadcom. These parts may not be generally available
in small quantities for embedded development.
The GEC block requires an opaque firmware blob to function. The
GEC blob is controlled by AMD and Broadcom and is not available
from coreboot.org.
This change removes GEC support from AMD Parmer and AMD Thatcher
mainboards since these boards do not have the Broadcom PHY.
AMD has requested that the GEC be hidden for Hudson FCH since
the PHY parts are not generally available. This Kconfig option
can make it appear that this is a viable and supported way to
add Ethernet to an embedded board. It is possible to use the
Hudson GEC block with other PHYs, but this requires development
of a custom GEC blob and a custom Ethernet driver. A custom GEC
blob has been developed for a Micrel PHY, but there is no
accompanying driver.
Change-Id: I7a7bf4d41e453390ecf987c9c45ef2434fc1f1a3
Signed-off-by: Bruce Griffith <Bruce.Griffith@se-eng.com>
Reviewed-on: http://review.coreboot.org/3127
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Tested-by: build bot (Jenkins)
Reviewed-by: Jens Rottmann <JRottmann@LiPPERTembedded.de>
Reviewed-by: Martin Roth <martin.roth@se-eng.com>
It is useful to be able to lock out certain address ranges,
NULL being the most important example.
void mmu_disable_range(unsigned long start_mb, unsigned long size_mb)
will allow us to lock out selected virtual addresses on MiB boundaries.
As in other ARM mmu functions, the addresses and quantities are in units
of MiB.
Change-Id: If516ce955ee2d12c5a409f25acbb5a4b424f699b
Signed-off-by: Ronald G. Minnich <rminnich@gmail.com>
Reviewed-on: http://review.coreboot.org/3160
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Tested-by: build bot (Jenkins)
Reviewed-by: David Hendricks <dhendrix@chromium.org>
It's fine to always start timer even in suspend/resume mode, so we can
move the timer_start() back to the very beginning of boot procedure.
That provides more precise boot time information.
With that timer change, the wake up state test procedure can be simplified.
Verified by building and booting firmware image on Google/Snow successfully,
and then suspend-resume without problem (suspend_stress_test).
Change-Id: I0d739650dbff4eb3a75acbbf1e4356f2569b487d
Signed-off-by: Hung-Te Lin <hungte@chromium.org>
Reviewed-on: http://review.coreboot.org/3151
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
This adds an inline wrapper for the TLBIMVAA instruction (invalidate
unified TLB by MVA, all address space identifiers).
Change-Id: Ibcd289ecedaba8586ade26e36c177ff1fcaf91d3
Signed-off-by: David Hendricks <dhendrix@chromium.org>
Reviewed-on: http://review.coreboot.org/3161
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
Tested-by: build bot (Jenkins)
The firmware media source (SPI1) is already initialized by Exynos iROM.
There is no need to do it again.
Verified by building and booting Google/Snow successfully.
Change-Id: I89390506aa825397c0d7e52ad7503f1cb808f7db
Signed-off-by: Hung-Te Lin <hungte@chromium.org>
Reviewed-on: http://review.coreboot.org/3147
Tested-by: build bot (Jenkins)
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
When TIMER_QUEUE is configured on call the timer callbacks on
entry into a state but before its entry callbacks. In addition
provide a barrier to the following states so that timers are drained
before proceeding. This allows for blocking state traversal for key
components of boot.
BS_OS_RESUME
BS_WRITE_TABLES
BS_PAYLOAD_LOAD
BS_PAYLOAD_BOOT
Future functionality consists of evaluating the timer callbacks within
the device tree. One example is dev_initialize() as that seems state
seems to take 90% of the boot time. The timer callbacks could then be
ran in a more granular manner.
Change-Id: Idb549ea17c5ec38eb57b4f6f366a1c2183f4a6dd
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/3159
Tested-by: build bot (Jenkins)
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
A timer queue provides the mechanism for calling functions
in the future by way of a callback. It utilizes the MONOTONIC_TIMER
to track time through the boot. The implementation is a min-heap
for keeping track of the next-to-expire callback.
Change-Id: Ia56bab8444cd6177b051752342f53b53d5f6afc1
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/3158
Tested-by: build bot (Jenkins)
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
When the MONOTONIC_TIMER is available track the entry, run, and exit
times for each state. It should be noted that the times for states that
vector to OS or a payload do not have their times reported.
Change-Id: I6af23fe011609e0b1e019f35ee40f1fbebd59c9d
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/3156
Tested-by: build bot (Jenkins)
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
Implement the timer_monotonic_get() using the TSC.
Change-Id: I5118da6fb9bccc75d2ce012317612e0ab20a2cac
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/3155
Tested-by: build bot (Jenkins)
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
Implement the timer_monotonic_get() functionality based off of
the local apic timer.
Change-Id: I1aa1ff64d15a3056d6abd1372be13da682c5ee2e
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/3154
Tested-by: build bot (Jenkins)
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
Haswell ULT devices have a 24MHz package-level counter. Use
this counter to provide a timer_monotonic_get() implementation.
Change-Id: Ic79843fcbfbbb6462ee5ebd12b39502307750dbb
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/3153
Tested-by: build bot (Jenkins)
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
The notion of a monotonic timer is introduced. Along with it
are helper functions and other types for comparing times. This
is just the framework where it is the responsibility of the
chipset/board to provide the implementation of timer_monotonic_get().
The reason structs are used instead of native types is to allow
for future changes to the data structure without chaning all the
call sites.
Change-Id: Ie56b9ab9dedb0da69dea86ef87ca744004eb1ae3
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/3152
Tested-by: build bot (Jenkins)
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
The notion of loading a payload in the current boot state
machine isn't actually loading the payload. The reason is
that cbfs is just walked to find the payload. The actual
loading and booting were occuring in selfboot(). Change this
balance so that loading occurs in one function and actual
booting happens in another. This allows for ample opportunity
to delay work until just before booting.
Change-Id: Ic91ed6050fc5d8bb90c8c33a44eea3b1ec84e32d
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/3139
Tested-by: build bot (Jenkins)
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
On x86 systems there is a concept of cachings the ROM. However,
the typical policy is that the boot cpu is the only one with
it enabled. In order to ensure the MTRRs are the same across cores
the rom cache needs to be disabled prior to OS resume or boot handoff.
Therefore, utilize the boot state callbacks to schedule the disabling
of the ROM cache at the ramstage exit points.
Change-Id: I4da5886d9f1cf4c6af2f09bb909f0d0f0faa4e62
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/3138
Tested-by: build bot (Jenkins)
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
The cbmem_post_handling() function was implemented by 2
chipsets in order to save memory configuration in flash. Convert
both of these chipsets to use the boot state machine callbacks
to perform the saving of the memory configuration.
Change-Id: I697e5c946281b85a71d8533437802d7913135af3
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/3137
Tested-by: build bot (Jenkins)
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
There were previously 2 functions, init_cbmem_pre_device() and
init_cbmem_post_device(), where the 2 cbmem implementations
implemented one or the other. These 2 functions are no longer
needed to be called in the boot flow once the boot state callbacks
are utilized.
Change-Id: Ida71f1187bdcc640ae600705ddb3517e1410a80d
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/3136
Tested-by: build bot (Jenkins)
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
Utilize the static boot state callback scheduling to initialize
and tear down the coverage infrastructure at the appropriate points.
The coverage initialization is performed at BS_PRE_DEVICE which is the
earliest point a callback can be called. The tear down occurs at the
2 exit points of ramstage: OS resume and payload boot.
Change-Id: Ie5ee51268e1f473f98fa517710a266e38dc01b6d
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/3135
Tested-by: build bot (Jenkins)
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
It's helpful to provide a distinct state that affirmatively
describes that OS resume will occur. The previous code included
the check and the actual resuming in one function. Because of this
grouping one had to annotate the innards of the ACPI resume
path to perform specific actions before OS resume. By providing
a distinct state in the boot state machine the necessary actions
can be scheduled accordingly without modifying the ACPI code.
Change-Id: I8b00aacaf820cbfbb21cb851c422a143371878bd
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/3134
Tested-by: build bot (Jenkins)
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
Many of the boot state callbacks can be scheduled at compile time.
Therefore, provide a way for a compilation unit to inform the
boot state machine when its callbacks should be called. Each C
module can export the callbacks and their scheduling requirements
without changing the shared boot flow code.
Change-Id: Ibc4cea4bd5ad45b2149c2d4aa91cbea652ed93ed
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/3133
Tested-by: build bot (Jenkins)
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
The boot flow currently has a fixed ordering. The ordering
is dictated by the device tree and on x86 the PCI device ordering
for when actions are performed. Many of the new machines and
configurations have dependencies that do not follow the device
ordering.
In order to be more flexible the concept of a boot state machine
is introduced. At the boundaries (entry and exit) of each state there
is opportunity to run callbacks. This ability allows one to schedule
actions to be performed without adding board-specific code to
the shared boot flow.
Change-Id: I757f406c97445f6d9b69c003bb9610b16b132aa6
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/3132
Tested-by: build bot (Jenkins)
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
While debugging a crash it was discovered that ld was inserting
address space for sections that were empty depending on section
address boundaries. This led to the assumption breaking down that
on-disk payload (code/data bits) was contiguous with the address
space. When that assumption breaks down relocation updates change
the wrong memory. Fix this by making the rmodule.ld linker script
put all code/data bits into a payload section.
Change-Id: Ib5df7941bbd64662090136e49d15a570a1c3e041
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/3149
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Tested-by: build bot (Jenkins)
Paul Menzel