A dedicated pci device driver required for LPC devices as the legacy
IO range need to be included to avoid IO resource confilict. Blindly
set to 0~0x1000 to also avoid the IO resource of COMA/COMB/LPT/FDD
and LPC.Without this driver system will have assertion on load
RTC DXE driver in UEFI payloads.
Change-Id: Icc462c159c2cf39cc1030d55acee79e73a6bfb35
Signed-off-by: Lance Zhao <lijian.zhao@intel.com>
Reviewed-on: https://review.coreboot.org/13356
Tested-by: build bot (Jenkins)
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-by: Martin Roth <martinroth@google.com>
ACPI MADT tables required to describe the multiprocessor interrupt
routing. Apollolake SOC also have the interrupt override table like
other x86 silicons.
Change-Id: I85976e227963c950aad4476d68581b96e1090559
Signed-off-by: Lance Zhao <lijian.zhao@intel.com>
Reviewed-on: https://review.coreboot.org/13373
Tested-by: build bot (Jenkins)
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Two of the MCT data structures passed as substructures to ramstage were
not packed, and additionally no alignment was specified. On at least
SP5100-based platforms, specifying packed with no alignment caused boot
failure dependent on the exact compiled binary layout (LPC hang).
Specifying the alignment and packing the remaining structures appears to
have resolved the remaining LPC hang issues on the KGPE-D16. Note that
packing the remaining structures alone was not sufficient to eliminate
the hang, however removing the packed attribute entirely (during debugging)
did resolve the hang at the expense of potential problems in ramstage.
Change-Id: If3a7509ed438870d4d05caaaaa091e1c47bf9b97
Signed-off-by: Timothy Pearson <tpearson@raptorengineeringinc.com>
Reviewed-on: https://review.coreboot.org/14303
Reviewed-by: Martin Roth <martinroth@google.com>
Tested-by: build bot (Jenkins)
Tested-by: Raptor Engineering Automated Test Stand <noreply@raptorengineeringinc.com>
This enables CACHE_MRC_SETTINGS by default as well selects
timer configuration.
Change-Id: I0248001892ef763c39097848b5adc8c1befed1f0
Signed-off-by: Hannah Williams <hannah.williams@intel.com>
Signed-off-by: Andrey Petrov <andrey.petrov@intel.com>
Reviewed-on: https://review.coreboot.org/14252
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
The SPI controller needs to be set up on devices such as the SP5100
before it can be accessed to write MCT backup data. Move the backup
data write after PCI configuration has been completed.
Change-Id: Ibcf31755242ac058407a422ce8aa33d6b0b293c7
Signed-off-by: Timothy Pearson <tpearson@raptorengineeringinc.com>
Reviewed-on: https://review.coreboot.org/14305
Tested-by: build bot (Jenkins)
Tested-by: Raptor Engineering Automated Test Stand <noreply@raptorengineeringinc.com>
Reviewed-by: Martin Roth <martinroth@google.com>
ACPI MCFG table is required for OS to support Enhanced
Configuration Space Access.Apollolake will only support
1 PCI Segment Group, so all the pci bus number from 0
to 0xff will belong to that group.
Change-Id: I3a680eb9c83290cd531159d7e796382a132cd283
Signed-off-by: Lance Zhao <lijian.zhao@intel.com>
Reviewed-on: https://review.coreboot.org/13375
Tested-by: build bot (Jenkins)
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Implement flash read, write, and erase functionality using the
hardware sequencing capabilities of the SOC. Due to changes in
hardware requirements, the flash chip must be probed differently
than on previous platforms (details explained in comments).
Note that this is a minimal implementation, and does not provide all
the bells and whistles.
Change-Id: I6dcc3bc36dfce61927d126d231a16d485acb1bdc
Signed-off-by: Alexandru Gagniuc <alexandrux.gagniuc@intel.com>
Signed-off-by: Andrey Petrov <andrey.petrov@intel.com>
Reviewed-on: https://review.coreboot.org/14246
Tested-by: build bot (Jenkins)
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
The delay_tsc.c compilation unit used the C preprocessor
to conditionally compile different code paths. Instead of
guarding large blocks of code allow the compiler to optimize
out unreachable code.
Change-Id: I660c21d6f4099b0d7aefa84b14f1e68d6fd732c3
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: https://review.coreboot.org/14302
Tested-by: build bot (Jenkins)
Reviewed-by: Furquan Shaikh <furquan@google.com>
Reviewed-by: Andrey Petrov <andrey.petrov@intel.com>
The delay_tsc.c code took different paths depending
__PRE_RAM__ being defined or not. Also, timer_monotonic_get()
was only compiled in a !__PRE_RAM__ environment. Clean up
the code paths by employing CAR_GLOBAL for the global state
which allows the same code to be used in all stages.
Lastly, handle apollolake fallout now that init_timer() is
not needed in placeholders.c.
Change-Id: Ia769fa71e2c9d8b11201a3896d117097f2cb7c56
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: https://review.coreboot.org/14301
Tested-by: build bot (Jenkins)
Reviewed-by: Furquan Shaikh <furquan@google.com>
Reviewed-by: Andrey Petrov <andrey.petrov@intel.com>
The current code in delay_tsc.c uses globals and is heavily
guarded by a lot of preprocessor macros. In order to remove
__PRE_RAM__ constraints one needs to use CAR_GLOBAL for the
global variables. Therefore, abstract away direct access to
the globals such that CAR_GLOBAL can be easily employed.
Change-Id: I3350d1a762120476926c8d9f5f5a7aba138daf5f
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: https://review.coreboot.org/14300
Tested-by: build bot (Jenkins)
Reviewed-by: Furquan Shaikh <furquan@google.com>
Reviewed-by: Andrey Petrov <andrey.petrov@intel.com>
It's not selected by any path so it's a dead option with
associated dead code. Remove the config option as well as
the code paths that were never used any longer.
Change-Id: Ie536eee54e5c63bd90192f413c69e0dd2fea9171
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: https://review.coreboot.org/14299
Tested-by: build bot (Jenkins)
Reviewed-by: Furquan Shaikh <furquan@google.com>
Reviewed-by: Andrey Petrov <andrey.petrov@intel.com>
Reviewed-by: Myles Watson <mylesgw@gmail.com>
On certain versions of /bin/sh the following sequence
causes problems.
'$CC --version | grep clang &>/dev/null && ...'
The above is a bashish for 2>&1 >/dev/null. However, buildgcc
is interpeted by /bin/sh which doesn't necessarily mean bash.
On dash it's effectively forking grep off into the background
and always evaluating an empty statement to /dev/null while
unconditionally running whatever follows the &&.
Change-Id: Ie3a2ebb12226434d50a7b2a7e254c8b80ae4c46b
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: https://review.coreboot.org/14281
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
The LEDs on the beaglebone are connected to GPIOs called USR0-USR3. This
change adds some functions to make it easy to set their value and clear
what the calling code is trying to do.
Change-Id: I0bb83bbc2e195ce1a0104afcd120089efaa22916
Signed-off-by: Gabe Black <gabeblack@chromium.org>
Reviewed-on: https://review.coreboot.org/3943
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
Add code for manipulating the GPIOs on the am335x. The API is patterned after
the one used for the Exynos SOCs.
Change-Id: I275317304bd0682f348f72f1c77ed5613065af3f
Signed-off-by: Gabe Black <gabeblack@chromium.org>
Reviewed-on: https://review.coreboot.org/3942
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
To avoid having to read/write raw addresses with magic constants,
this change adds data structures which represent the clock module
registers and some constants for how the clock module is used
currently.
Change-Id: I955dae39bbdabccf048a086e706a48c58f620ad4
Signed-off-by: Gabe Black <gabeblack@chromium.org>
Reviewed-on: https://review.coreboot.org/3941
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
Always use MRC cache if possible.
Added a CRC16 array to make sure the DIMMs haven't been replaced.
In case one of the CRC's doesn't match, start normal RAM training.
Use new fallback in case of broken mrc cache.
Test system:
* Gigabyte GA-B75M-D3H
* Intel Pentium CPU G2130
Test result:
The system boots a lot faster using the MRC cache.
On swapping DIMMs the CRC16 doesn't match and normal ram training
is started.
Change-Id: Ib48fe8380446846df17d37b22968f7d4fd6b9b13
Signed-off-by: Patrick Rudolph <siro@das-labor.org>
Reviewed-on: https://review.coreboot.org/14172
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
On s3 wakeup h8_enable is called which resets the (audio) volume. But the
volume should be the same as before the s3 state. In particular, userland
programs (e.g. pulseaudio) may be out of sync, if the volume can be changed
by hardware buttons also emitting acpi events. Hence, do not reset the
volume on s3 wakeup.
Tested on a Lenovo ThinkPad X220.
Change-Id: I2af08dea1a3f14a40734d67d372e845cc18c5e09
Signed-off-by: Christopher Spinrath <christopher.spinrath@rwth-aachen.de>
Reviewed-on: https://review.coreboot.org/14183
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Alexander Couzens <lynxis@fe80.eu>
There are more modules in a category than categories. Moving the clock
down leaves more space for the list of modules.
Change-Id: I536dafe32e1abb1995c8a1942d70e0d90b905612
Signed-off-by: Jonathan Neuschäfer <j.neuschaefer@gmx.net>
Reviewed-on: https://review.coreboot.org/14255
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Martin Roth <martinroth@google.com>
Add a script to help us verify that our lint tests are working.
This isn't finished, because it should test all of the failure modes.
Some of the tests, 008-kconfig in particular have a lot of ways
that they can fail.
Currently the Kconfig test is triggered by removing the board
name file in test 006. This removes the only place the config
option for that board name is located.
Change-Id: If01c6daf1c99d097a19995b4befae90a3b5db2d6
Signed-off-by: Martin Roth <martinroth@google.com>
Reviewed-on: https://review.coreboot.org/14198
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
The removable DIMM SPD data wasn't read.
As a result the system only uses the 2GB onboard memory and
the GNU Linux kernel paniced in acpi_ds_build_internal_package_obj.
Read the SPD and allow native raminit and MRC blob to use the
removable DIMM.
The system is able to use the removable dimm and the kernel panic
is gone.
Change-Id: I30eed747f924cb0029de55d2ab85c5a94075dc1b
Signed-off-by: Patrick Rudolph <siro@das-labor.org>
Tested-by: Idwer Vollering <vidwer@gmail.com>
Reviewed-on: https://review.coreboot.org/14278
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
libgcc fails to compile on a number of platforms when a
non-GNU sed is used.
This patch has been verified by building the MIPS reference
toolchain on OS X.
Change-Id: Ia1c18ea4359de7707ac2e2640f1b8f107c47cd8c
Signed-off-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
Reviewed-on: https://review.coreboot.org/14275
Tested-by: build bot (Jenkins)
Reviewed-by: Idwer Vollering <vidwer@gmail.com>
On certain versions of /bin/sh assigning variables with spaces
unquoted leads to failures. Therefore, quote variables that
are known to be passed in that have spaces.
Change-Id: I007c56c3bfb8183bb4b16cf0591f6aa508fd105d
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: https://review.coreboot.org/14280
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
Reviewed-by: Martin Roth <martinroth@google.com>
This used to build, but will not with newer toolchains.
Change-Id: I0f397839eb85977ba18328b0e32040b15a6c3b0f
Signed-off-by: Ronald G. Minnich <rminnich@gmail.com>
Reviewed-on: https://review.coreboot.org/14296
Tested-by: build bot (Jenkins)
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
This resolves error messages of the form:
ERROR: device PNP: 002e.6 index 98 has no mask.
Change-Id: I6a368b902d051c8da6f74cbde54f5d12a3e52c2f
Signed-off-by: Timothy Pearson <tpearson@raptorengineeringinc.com>
Reviewed-on: https://review.coreboot.org/14272
Tested-by: build bot (Jenkins)
Tested-by: Raptor Engineering Automated Test Stand <noreply@raptorengineeringinc.com>
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
The default nvm_mmio_to_flash_offset() implementation used by NVM code
in intel/common does not work on apollolake. As a result, provide the
correct override.
Change-Id: I01a94f90dfdd33586a4aac5c05dd8c73e8804437
Signed-off-by: Alexandru Gagniuc <alexandrux.gagniuc@intel.com>
Signed-off-by: Andrey Petrov <andrey.petrov@intel.com>
Reviewed-on: https://review.coreboot.org/14248
Tested-by: build bot (Jenkins)
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
On apollolake, the flash is memory-mapped differently, and the default
MMIO to flash calculation does not produce correct results. While the
long-term solution is to rewrite the NVM functionality to keep the
flash offset as part of its context, as a temporary measure, allow
overriding the to_flash_offset() function by declaring it weak.
Change-Id: Ic54baeba2441a08cfe1a47e235747797f6efb59b
Signed-off-by: Alexandru Gagniuc <alexandrux.gagniuc@intel.com>
Signed-off-by: Andrey Petrov <andrey.petrov@intel.com>
Reviewed-on: https://review.coreboot.org/14247
Tested-by: build bot (Jenkins)
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
This reverts commit f961becc43.
On studying the BKDG more closely this is not the correct place
to enable DIMM parity. Further patches to clarify the parity
setup process on Family 15h are forthcoming.
Change-Id: I5a3a4f1621e3048f9dfc159709410be9de6ebecd
Reviewed-on: https://review.coreboot.org/14271
Tested-by: build bot (Jenkins)
Tested-by: Raptor Engineering Automated Test Stand <noreply@raptorengineeringinc.com>
Reviewed-by: Martin Roth <martinroth@google.com>
The sync flood reset fix in Change-Id: I62d897010a8120aa14b4cb8d096bc4f2edc5f248
and related changes have made it possible to move the sync flood enable statements
back into romstage.
Change-Id: I5a3a4f1621e3048f9dfc159709410be9de6ebece
Signed-off-by: Timothy Pearson <tpearson@raptorengineeringinc.com>
Reviewed-on: https://review.coreboot.org/14270
Tested-by: build bot (Jenkins)
Tested-by: Raptor Engineering Automated Test Stand <noreply@raptorengineeringinc.com>
Reviewed-by: Martin Roth <martinroth@google.com>
When a fatal error and subsequent sync flood / reset occurs,
the MCA status registers may contain valuable information on
the cause of the fatal error. Add functions to report MCEs and
reset the MCA status registers early in the boot process.
Change-Id: Icde1051ac22f93688de1330f5e2c9ce28b14b59a
Signed-off-by: Timothy Pearson <tpearson@raptorengineeringinc.com>
Reviewed-on: https://review.coreboot.org/14265
Tested-by: build bot (Jenkins)
Tested-by: Raptor Engineering Automated Test Stand <noreply@raptorengineeringinc.com>
Reviewed-by: Martin Roth <martinroth@google.com>
The SB700 family has the ability to report the last reset
reason. This is useful in the context of handling MCEs
and recovering from fatal errors / sync floods.
Add a function to retrieve the last reset flags.
Change-Id: I754cb25e47bd9c1e4a29ecb6cb18017d1b7c3dc4
Signed-off-by: Timothy Pearson <tpearson@raptorengineeringinc.com>
Reviewed-on: https://review.coreboot.org/14263
Tested-by: Raptor Engineering Automated Test Stand <noreply@raptorengineeringinc.com>
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Martin Roth <martinroth@google.com>
Certain AMD platforms, such as those using the SP5100 southbridge,
contain a very poorly documented bug related to LPC ROM access,
which is triggered by repeated (hundreds or more) rapid calls to
get_option(). This bug manifests as a complete system deadlock
in ramstage device configuration, requiring standby power to be
removed from the system to release the deadlock.
Cache the platform ECC status to avoid repeated calls to get_option()
in the lane count detection logic.
Change-Id: I8b48c523218ccc8c113319957d6eca2d15e1070f
Signed-off-by: Timothy Pearson <tpearson@raptorengineeringinc.com>
Reviewed-on: https://review.coreboot.org/14273
Tested-by: build bot (Jenkins)
Tested-by: Raptor Engineering Automated Test Stand <noreply@raptorengineeringinc.com>
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Martin Roth <martinroth@google.com>
This adds boot mode constants. They match EDK2 found in PiBootMode.h
constants but are part of FSP2.0 spec.
Change-Id: I16ee90ff372d252ddc042ca89c1e5912ab041616
Signed-off-by: Andrey Petrov <andrey.petrov@intel.com>
Reviewed-on: https://review.coreboot.org/14249
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Tested-by: build bot (Jenkins)
This reverts commit 272a1f05b9.
In Chrome OS this command's usage was dropped in favor of another
solution. As it's not used drop the support for it.
Change-Id: I58b51446d3a8b5fed7fc391025225fbe38ffc007
Signed-off-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: https://review.coreboot.org/14261
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Martin Roth <martinroth@google.com>
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
Upcoming designs are based on similar SOCs, this patch moves code
which can be reused into a common directory under soc/rockchip.
Changing spi.h to include stdder.h, as this is were check_member() is
defined, this becomes necessary later when the new SOC code is added.
Renaming UART driver private functions not to be bound to any
particular SOC.
BUG=none
BRANCH=none
TEST=the refactored code works fine on the new platform (with the rest
of the patches applied).
Change-Id: I39a505aecda8849daa58a8eca0e44a5243664423
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: f63f2582042ac115481207ddf329ea2e3260e55e
Original-Change-Id: I3a1139305354d460492b25a45f3da315a9a0b49e
Original-Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/335408
Original-Reviewed-by: Julius Werner <jwerner@chromium.org>
Original-Reviewed-by: Patrick Georgi <pgeorgi@chromium.org>
Reviewed-on: https://review.coreboot.org/14235
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
Our EDID code had always been aligning the framebuffer's
bytes_per_line (and x_resolution dependent on that) to 64. It turns out
that this is a controller-dependent parameter that seems to only really
be necessary for Intel chipsets, and commit 6911219cc (edid: Add helper
function to calculate bits-per-pixel dependent values) probably actually
broke this for some other controllers by applying the alignment too
widely.
This patch makes it explicitly configurable and depends the default on
ARCH_X86 (which seems to be the simplest and least intrusive way to make
it fit most cases for now... boards where this doesn't apply can still
override it manually by calling edid_set_framebuffer_bits_per_pixel()
again).
Change-Id: I1c565a72826fc5ddfbb1ae4a5db5e9063b761455
Signed-off-by: Julius Werner <jwerner@chromium.org>
Reviewed-on: https://review.coreboot.org/14267
Tested-by: build bot (Jenkins)
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
The logic to enable reset on sync flood per RPR guidelines
somehow ended up guarded on the SATA AHCI setup. Unconditionally
enable reset on sync flood per the RPR.
Change-Id: I62d897010a8120aa14b4cb8d096bc4f2edc5f248
Signed-off-by: Timothy Pearson <tpearson@raptorengineeringinc.com>
Reviewed-on: https://review.coreboot.org/14260
Tested-by: build bot (Jenkins)
Tested-by: Raptor Engineering Automated Test Stand <noreply@raptorengineeringinc.com>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Felix Held <felix-coreboot@felixheld.de>
Otherwise, on OS X, some architectures will fail
to build libgcc (verified for ARM toolchain).
Change-Id: I8b58e0582596ad39cad92e9d478158c46a96a26e
Signed-off-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
Reviewed-on: https://review.coreboot.org/14256
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
Most cross compilers fail to compile on systems with Clang being the
default compiler (OS X and some BSDs). Clang dislikes some of GCC's
autogenerated code. We also missed switching CFLAGS to CXXFLAGS when GCC
switched to C++ compilation per default.
Change-Id: I87caa1a15982c431048aa79748ea7ef655a9a3a1
Signed-off-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
Reviewed-on: https://review.coreboot.org/14232
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martinroth@google.com>
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
Since r is a pointer, memset(r, 0, sizeof(r)) would only zero the first
4 (or 8) bytes of the newly allocated struct align_region_t.
An alternative to this patch would be to use calloc, or introduce a new
zalloc (zeroed allocation; a single-element calloc) and use that.
Change-Id: Ic3e3487ce749eeebf6c4836e62b8a305ad766e7e
Found-by: Coverity (ID 1291160)
Signed-off-by: Jonathan Neuschäfer <j.neuschaefer@gmx.net>
Reviewed-on: https://review.coreboot.org/14244
Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
power_on_after_fail=Enable in cmos.default leads to wake on AC behaviour
on mobile systems. Therefore set cmos.default entry to "Disable" in order
to improve user experience.
Change-Id: I977a4e6bc028c8c4c7fc1c2f5fdd74a59e951c60
Signed-off-by: Philipp Deppenwiese <zaolin@das-labor.org>
Reviewed-on: https://review.coreboot.org/13884
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Alexander Couzens <lynxis@fe80.eu>
Zhao, Lijian