This patch removes the wbinvd call preceding CR0.CD setting in
disable_cache() to improve the boot time performances. According to
some experimental measurements, the wbinvd execution takes between 1.6
up and 6 milliseconds to complete so it is preferable to call it only
when necessary.
According to Intel Software Developer Manual Vol 3.A - 12.5.3
Preventing Caching section there is no need to flush and invalidate
the cache before settings CR0.CD. The documented sequence consists in
setting CR0.CD and then call wbinvd.
We also could not find any extra requirements in the AMD64
Architecture Programmer’s Manual - Volume 2 - Memory System chapter.
This extra wbinvd in coreboot disable_cache() function does not seem
documented and looking into the history of the project got us all the
way back to original commit 8ca8d7665d67 ("- Initial checkin of the
freebios2 tree") from April 2003.
Even the original disable_cache() implementation (see below) is a bit
curious as the comment list two actions:
1. Disable cache cover by line 74, 75 and 77
2. Write back the cache and flush TLB - Line 78
But it does not provide any explanation for the wbinvd call line 76.
68 static inline void disable_cache(void)
69 {
70 unsigned int tmp;
71 /* Disable cache */
72 /* Write back the cache and flush TLB */
73 asm volatile (
74 "movl %%cr0, %0\n\t"
75 "orl $0x40000000, %0\n\t"
76 "wbinvd\n\t"
77 "movl %0, %%cr0\n\t"
78 "wbinvd\n\t"
79 :"=r" (tmp)
80 ::"memory");
81 }
BUG=b/260455826
TEST=Successful boot on Skolas and Rex board
Change-Id: I08c6486dc93c4d70cadc22a760d1b7e536e85bfa
Signed-off-by: Jeremy Compostella <jeremy.compostella@intel.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/75474
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Subrata Banik <subratabanik@google.com>
Reviewed-by: Himanshu Sahdev <himanshu.sahdev@intel.com>
Improve boot time performances by replacing the wbinvd instruction
with multiple clflush to ensure that the SIPI data is written back to
RAM.
According to some experimental measurements, the wbinvd execution
takes between 1.6 up and 6 milliseconds to complete. In the case of
the SIPI data, wbinvd unnecessarily flushes and invalidates the entire
cache. Indeed, the SIPI module is quite small (about 400 bytes) and
cflush'ing the associated cache lines is almost instantaneous,
typically less than 100 microseconds.
BUG=b/260455826
TEST=Successful boot on Skolas and Rex board
Change-Id: I0e00db8eaa6a3cb41bec3422572c8f2a9bec4057
Signed-off-by: Jeremy Compostella <jeremy.compostella@intel.com>
Suggested-by: Erin Park <erin.park@intel.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/75391
Reviewed-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Arthur Heymans <arthur@aheymans.xyz>
When cbmem is initialized in romstage and postcar placed in the stage
cache + cbmem where it is run, the assumption is made that these are
all in UC memory such that calling INVD in postcar is OK.
For performance reasons (e.g. postcar decompression) it is desirable
to cache cbmem and the stage cache during romstage.
Another reason is that AGESA sets up MTRR during romstage to cache all
dram, which is currently worked around by using additional MTRR's to
make that UC.
TESTED on asus/p5ql-em, up/squared on both regular and S3 resume
bootpath. Sometimes there are minimal performance improvements
when cbmem is cached (few ms).
Change-Id: I7ff2a57aee620908b71829457ea0f5a0c410ec5b
Signed-off-by: Arthur Heymans <arthur@aheymans.xyz>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/37196
Reviewed-by: Lean Sheng Tan <sheng.tan@9elements.com>
Reviewed-by: Kapil Porwal <kapilporwal@google.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Selecting CPU_X86_CACHE_HELPER only added the x86_enable_cache wrapper
function around enable_cache which additionally wrote a POST code to
port 0x80 and printed a message to the console. This function was only
called during multi-processor initialization in ramstage via the init
function pointer in the CPU's device operations struct and was run on
all cores, so the message on the console was printed once per CPU core.
This patch replaces all x86_enable_cache calls by calls to enable_cache
and removes the wrapper function and the Kconfig symbol
CPU_X86_CACHE_HELPER which was used to only add this when the
corresponding CPUs used the x86_enable_cache wrapper function.
Signed-off-by: Felix Held <felix-coreboot@felixheld.de>
Suggested-by: Angel Pons <th3fanbus@gmail.com>
Change-Id: I5866b6bf014821ff9e3a48052a5eaf69319b003a
Reviewed-on: https://review.coreboot.org/c/coreboot/+/58579
Reviewed-by: Arthur Heymans <arthur@aheymans.xyz>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
INVD is called below so if postcar is running in a cached environment
it needs to happen.
NOTE: postcar cannot execute in a cached environment if clflush is not
supported!
Change-Id: I37681ee1f1d2ae5f9dd824b5baf7b23b2883b1dc
Signed-off-by: Arthur Heymans <arthur@aheymans.xyz>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/37212
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-by: Patrick Rudolph <siro@das-labor.org>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Stefan thinks they don't add value.
Command used:
sed -i -e '/file is part of /d' $(git grep "file is part of " |egrep ":( */\*.*\*/\$|#|;#|-- | *\* )" | cut -d: -f1 |grep -v crossgcc |grep -v gcov | grep -v /elf.h |grep -v nvramtool)
The exceptions are for:
- crossgcc (patch file)
- gcov (imported from gcc)
- elf.h (imported from GNU's libc)
- nvramtool (more complicated header)
The removed lines are:
- fmt.Fprintln(f, "/* This file is part of the coreboot project. */")
-# This file is part of a set of unofficial pre-commit hooks available
-/* This file is part of coreboot */
-# This file is part of msrtool.
-/* This file is part of msrtool. */
- * This file is part of ncurses, designed to be appended after curses.h.in
-/* This file is part of pgtblgen. */
- * This file is part of the coreboot project.
- /* This file is part of the coreboot project. */
-# This file is part of the coreboot project.
-# This file is part of the coreboot project.
-## This file is part of the coreboot project.
--- This file is part of the coreboot project.
-/* This file is part of the coreboot project */
-/* This file is part of the coreboot project. */
-;## This file is part of the coreboot project.
-# This file is part of the coreboot project. It originated in the
- * This file is part of the coreinfo project.
-## This file is part of the coreinfo project.
- * This file is part of the depthcharge project.
-/* This file is part of the depthcharge project. */
-/* This file is part of the ectool project. */
- * This file is part of the GNU C Library.
- * This file is part of the libpayload project.
-## This file is part of the libpayload project.
-/* This file is part of the Linux kernel. */
-## This file is part of the superiotool project.
-/* This file is part of the superiotool project */
-/* This file is part of uio_usbdebug */
Change-Id: I82d872b3b337388c93d5f5bf704e9ee9e53ab3a9
Signed-off-by: Patrick Georgi <pgeorgi@google.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/41194
Reviewed-by: HAOUAS Elyes <ehaouas@noos.fr>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Done with sed and God Lines. Only done for C-like code for now.
Change-Id: I2fa3bad88bb5b068baa1cfc6bbcddaabb09da1c5
Signed-off-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/40053
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
They're listed in AUTHORS and often incorrect anyway, for example:
- What's a "Copyright $year-present"?
- Which incarnation of Google (Inc, LLC, ...) is the current
copyright holder?
- People sometimes have their editor auto-add themselves to files even
though they only deleted stuff
- Or they let the editor automatically update the copyright year,
because why not?
- Who is the copyright holder "The coreboot project Authors"?
- Or "Generated Code"?
Sidestep all these issues by simply not putting these notices in
individual files, let's list all copyright holders in AUTHORS instead
and use the git history to deal with the rest.
Change-Id: I89b10076e0f4a4b3acd59160fb7abe349b228321
Signed-off-by: Patrick Georgi <pgeorgi@google.com>
Reviewed-on: https://review.coreboot.org/c/coreboot/+/39611
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-by: Angel Pons <th3fanbus@gmail.com>
Reviewed-by: David Hendricks <david.hendricks@gmail.com>
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Its spreading copies got out of sync. And as it is not a standard header
but used in commonlib code, it belongs into commonlib. While we are at
it, always include it via GCC's `-include` switch.
Some Windows and BSD quirk handling went into the util copies. We always
guard from redefinitions now to prevent further issues.
Change-Id: I850414e6db1d799dce71ff2dc044e6a000ad2552
Signed-off-by: Nico Huber <nico.h@gmx.de>
Reviewed-on: https://review.coreboot.org/28927
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Add a __always_inline macro that wraps __attribute__((always_inline))
and replace current users with the macro, excluding files under
src/vendorcode.
Change-Id: Ic57e474c1d2ca7cc0405ac677869f78a28d3e529
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: https://review.coreboot.org/28587
Tested-by: build bot (Jenkins) <no-reply@coreboot.org>
Reviewed-by: Julius Werner <jwerner@google.com>
It encourages users from writing to the FSF without giving an address.
Linux also prefers to drop that and their checkpatch.pl (that we
imported) looks out for that.
This is the result of util/scripts/no-fsf-addresses.sh with no further
editing.
Change-Id: Ie96faea295fe001911d77dbc51e9a6789558fbd6
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Reviewed-on: http://review.coreboot.org/11888
Tested-by: build bot (Jenkins)
Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
As per discussion with lawyers[tm], it's not a good idea to
shorten the license header too much - not for legal reasons
but because there are tools that look for them, and giving
them a standard pattern simplifies things.
However, we got confirmation that we don't have to update
every file ever added to coreboot whenever the FSF gets a
new lease, but can drop the address instead.
util/kconfig is excluded because that's imported code that
we may want to synchronize every now and then.
$ find * -type f -exec sed -i "s:Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, *MA[, ]*02110-1301[, ]*USA:Foundation, Inc.:" {} +
$ find * -type f -exec sed -i "s:Foundation, Inc., 51 Franklin Street, Suite 500, Boston, MA 02110-1335, USA:Foundation, Inc.:" {} +
$ find * -type f -exec sed -i "s:Foundation, Inc., 59 Temple Place[-, ]*Suite 330, Boston, MA *02111-1307[, ]*USA:Foundation, Inc.:" {} +
$ find * -type f -exec sed -i "s:Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.:Foundation, Inc.:" {} +
$ find * -type f
-a \! -name \*.patch \
-a \! -name \*_shipped \
-a \! -name LICENSE_GPL \
-a \! -name LGPL.txt \
-a \! -name COPYING \
-a \! -name DISCLAIMER \
-exec sed -i "/Foundation, Inc./ N;s:Foundation, Inc.* USA\.* *:Foundation, Inc. :;s:Foundation, Inc. $:Foundation, Inc.:" {} +
Change-Id: Icc968a5a5f3a5df8d32b940f9cdb35350654bef9
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Reviewed-on: http://review.coreboot.org/9233
Tested-by: build bot (Jenkins)
Reviewed-by: Vladimir Serbinenko <phcoder@gmail.com>
The access to control registers were scattered about.
Provide a single header file to provide the correct
access function and definitions.
BUG=chrome-os-partner:22991
BRANCH=None
TEST=Built and booted using this infrastructure. Also objdump'd the
assembly to ensure consistency (objdump -d -r -S | grep xmm).
Change-Id: Iff7a043e4e5ba930a6a77f968f1fcc14784214e9
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: https://chromium-review.googlesource.com/172641
Reviewed-by: Stefan Reinauer <reinauer@google.com>
Reviewed-on: http://review.coreboot.org/4873
Tested-by: build bot (Jenkins)
Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Instead of opaque numbers like (1<<29), use
symbols like CR0_NoWriteThrough.
Change-Id: Id845e087fb472cfaf5f71beaf37fbf0d407880b5
Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Reviewed-on: http://review.coreboot.org/833
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
Root cause: After function STOP_CAR_AND_CPU disables cache as
ram, the cache as ram stack can no longer be used. Called
functions must be inlined to avoid stack usage. Also, the
compiler must keep local variables register based and not
allocated them from the stack. With gcc 4.5.0, some functions
declared as inline are not being inlined. This patch forces
these functions to always be inlined by adding the qualifier
__attribute__((always_inline)) to their declaration.
Signed-off-by: Scott Duplichan <scott@notabs.org>
Acked-by: Stefan Reinauer <stepan@coresystems.de>
Acked-by: Marc Jones <marcj303@gmail.com>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@5818 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1
Signed-off-by: Rudolf Marek <r.marek@assembler.cz>
Acked-by: Rudolf Marek <r.marek@assembler.cz>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@5564 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1
Signed-off-by: Rudolf Marek <r.marek@assembler.cz>
Acked-by: Rudolf Marek <r.marek@assembler.cz>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@5563 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1
also the GCC generated code and it looks all right. Please test on some
multicore CPU.
I added the "memory" clobber to read_cr0 / write_cr0 function as it is in Linux
Kernel. Seems that if this is missing, GCC is too smart and messes the order
of reads/writes to CR0 (not tested if really a problem here, but be safe for
future users of this function ;)
Signed-off-by: Rudolf Marek <r.marek@assembler.cz>
Acked-by: Stefan Reinauer <stepan@coresystems.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@5562 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1
while others dislike them being extra commits, let's clean them up once and
for all for the existing code. If it's ugly, let it only be ugly once :-)
Signed-off-by: Stefan Reinauer <stepan@coresystems.de>
Acked-by: Stefan Reinauer <stepan@coresystems.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@5507 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1
appropriate. Also, factor out post_code() for __PRE_RAM__ code and drop it from
some mainboards.
Signed-off-by: Stefan Reinauer <stepan@coresystems.de>
Acked-by: Stefan Reinauer <stepan@coresystems.de>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@5307 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1
__ROMCC__ now means "Don't use prototypes, since romcc doesn't support them."
__PRE_RAM__ means "Use simpler versions of functions, and no device tree."
There are probably some places where both are tested, but only one is needed.
Signed-off-by: Myles Watson <mylesgw@gmail.com>
Acked-by: Peter Stuge <peter@stuge.se>
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@4921 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1
Creator: Yinghai Lu <yhlu@tyan.com>
cache_as_ram for AMD and some intel
git-svn-id: svn://svn.coreboot.org/coreboot/trunk@1967 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1
