summaryrefslogtreecommitdiff
path: root/src
AgeCommit message (Collapse)Author
2013-02-11snow: fix high_tables_base calculationDavid Hendricks
It was off by a few orders of magnitude. D'oh. Change-Id: I9c8a3d5bd9ce261f914cfc7d05d86a1c61519b81 Signed-off-by: David Hendricks <dhendrix@chromium.org> Reviewed-on: http://review.coreboot.org/2355 Tested-by: build bot (Jenkins) Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
2013-02-11spi.h: Rename the spi.h to spi-generic.hZheng Bao
Since there are and will be other files in nb/sb folders, we change the general spi.h to a file name which is not easy to be duplicated. Change-Id: I6548a81206caa608369be044747bde31e2b08d1a Signed-off-by: Zheng Bao <zheng.bao@amd.com> Signed-off-by: zbao <fishbaozi@gmail.com> Reviewed-on: http://review.coreboot.org/2309 Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net> Tested-by: build bot (Jenkins) Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
2013-02-11Intel: Replace MSR 0xcd with MSR_FSB_FREQPatrick Georgi
And move the corresponding #define to speedstep.h Change-Id: I8c884b8ab9ba54e01cfed7647a59deafeac94f2d Signed-off-by: Patrick Georgi <patrick@georgi-clan.de> Reviewed-on: http://review.coreboot.org/2339 Tested-by: build bot (Jenkins) Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
2013-02-11AMD/Persimmon: Enable the 2nd COM portDave Frodin
The hardware is there, so turn it on. Change-Id: I40aff1e84a22a05599c62b9f0b20397df0a40b15 Signed-off-by: Dave Frodin <dave.frodin@se-eng.com> Reviewed-on: http://review.coreboot.org/2353 Tested-by: build bot (Jenkins) Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
2013-02-11Supermicro H8QGI: Substract 1 from MMCONF range limitKonstantin Aladyshev
MMCONF space is defined by two config parameters: MMCONF_BASE_ADDRESS (0xF800 0000) MMCONF_BUS_NUMBER (64) Coreboot allocates 1MB per bus, so MMCONF limit should be: 0xF800 0000 + 64*(0x0010 0000) - 1 = 0xFBFF FFFF Current code does not have (-1) component, this makes MMCONF limit equal 0xFC00 FFFF. Not 0xFC00 0000, because according to BKDG lower two bytes of MMIO limit always equal 0xFFFF: MMIOLimit = {MMIOLimitRegister[47:16], FFFFh}. Add (-1) to correct this issue. No functionality change has been experienced. The five times slower RAM speed compared to the proprietary vendor BIOS still remains. Change-Id: I2c6494c28bb8d36e54ceb2aa7d8d965b0103cbe9 Signed-off-by: Konstantin Aladyshev <aladyshev@nicevt.ru> Reviewed-on: http://review.coreboot.org/2193 Tested-by: build bot (Jenkins) Reviewed-by: Martin Roth <martin.roth@se-eng.com> Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
2013-02-11AMD SB900: fix warningsMartin Roth
Add a prototype to a .h file Remove an unused file (GppHp.c) from the build by deleting it from the makefile. I left the file since this is vendorcode. This is the code for PCIe hotplug. Inside GppHp.c, make functions not called from outside static. This obviously isn't important since the file isn't used, but for the sake of the cleanup I thought I'd go ahead with it... This was tested with the torpedo build. This fixes these warnings: src/vendorcode/amd/cimx/sb900/Dispatcher.c: In function 'LocateImage': src/vendorcode/amd/cimx/sb900/Dispatcher.c:193:38: warning: cast to pointer from integer of different size [-Wint-to-pointer-cast] src/vendorcode/amd/cimx/sb900/Usb.c:740:1: warning: no previous prototype for 'XhciA12Fix' [-Wmissing-prototypes] src/vendorcode/amd/cimx/sb900/GppHp.c:65:1: warning: no previous prototype for 'sbGppHotPlugSmiProcess' [-Wmissing-prototypes] src/vendorcode/amd/cimx/sb900/GppHp.c: In function 'sbGppHotPlugSmiProcess': src/vendorcode/amd/cimx/sb900/GppHp.c:76:5: warning: implicit declaration of function 'SbStall' [-Wimplicit-function-declaration] src/vendorcode/amd/cimx/sb900/GppHp.c: At top level: src/vendorcode/amd/cimx/sb900/GppHp.c:101:1: warning: no previous prototype for 'sbGppHotUnplugSmiProcess' [-Wmissing-prototypes] src/vendorcode/amd/cimx/sb900/GppHp.c:134:1: warning: no previous prototype for 'sbGppHotplugSmiCallback' [-Wmissing-prototypes] src/vendorcode/amd/cimx/sb900/GppHp.c: In function 'sbGppHotplugSmiCallback': src/vendorcode/amd/cimx/sb900/GppHp.c:158:5: warning: implicit declaration of function 'outPort80' [-Wimplicit-function-declaration] Change-Id: I5a1a20eeb81e1f4d59e3e3192f081e11d8506f56 Signed-off-by: Martin Roth <martin.roth@se-eng.com> Reviewed-on: http://review.coreboot.org/2349 Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net> Tested-by: build bot (Jenkins) Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
2013-02-11AMD S3: Add missing erasing flash sector for saving MTRR registerZheng Bao
It has worked up to now because the region is already erased the first time the board boots, and every additional boot the same data is being written over the old data.(by Dave Frodin) Change-Id: Id334c60668e31d23c1d552d0ace8eb6ae5513e6b Signed-off-by: Zheng Bao <zheng.bao@amd.com> Signed-off-by: zbao <fishbaozi@gmail.com> Reviewed-on: http://review.coreboot.org/2304 Tested-by: build bot (Jenkins) Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
2013-02-11AMD S3: Change the hardcoded data size to macros.Zheng Bao
Change-Id: Ieefc4213a6dee9c399826b1daa98bbf4bc10d881 Signed-off-by: Zheng Bao <zheng.bao@amd.com> Signed-off-by: zbao <fishbaozi@gmail.com> Reviewed-on: http://review.coreboot.org/2303 Tested-by: build bot (Jenkins) Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
2013-02-11snow: make build script erase 192KB instead of 128KBDavid Hendricks
This will make the build script wipe out more flash memory content. Our image is a bit bigger now that we're testing with payloads, so this is just added paranoia to prevent weird surprises caused by not flashing the full image. Change-Id: I31969922079e96886573d9d802266eb0052277cd Signed-off-by: David Hendricks <dhendrix@chromium.org> Reviewed-on: http://review.coreboot.org/2352 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-02-11armv7/exynos5250/snow: deprecate CONFIG_{RAMBASE,RAMTOP}David Hendricks
RAMBASE and RAMTOP are leftovers from the x86 port and do not apply the same way on ARM platforms. On x86 they refer to the low memory region where coreboot tables reside. However on ARM we don't have such a region which is architecturally defined. So instead we'll use the CPU-defined DRAM base address and the mainboard-defined DRAM size. This also has the pleasant side-effect of fixing the coreboot tables to not clobber ramstage code... Change-Id: I5548ecf05e82f9d9ecec8548fabdd99cc1e39c3b Signed-off-by: David Hendricks <dhendrix@chromium.org> Reviewed-on: http://review.coreboot.org/2351 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-02-10armv7/snow: Remove unused modules in bootblock and romstage.Hung-Te Lin
For Exynos/snow, cpu_info and power modules and also some parts of the GPIO API (which require timer and pwm modules) are not used in the current bootblock. Clock init only needs to be used if early console is enabled. Now our bootblock is 22420 bytes with early serial console and 11192 bytes without. Those include the 8KB BL1 region. Change-Id: I9c958dafb9cf522df0dcfbef373ce741aa162544 Signed-off-by: Hung-Te Lin <hungte@chromium.org> Signed-off-by: David Hendricks <dhendrix@chromium.org> Reviewed-on: http://review.coreboot.org/2322 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-02-10exynos/snow: move SPI GPIO setup to mainboard bootblock codeDavid Hendricks
This moves GPIO setup from chip-specific SPI code to mainboard- specific bootblock code. This makes exynos_spi_open a bit more generic so it can eventually be used for any SPI channel. This also benefits CBFS since the user can set media->context to to any set of SPI registers. Change-Id: I2bcb9de370df0a79353c14b4d021b471ddebfacd Signed-off-by: David Hendricks <dhendrix@chromium.org> Reviewed-on: http://review.coreboot.org/2347 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-02-10exynos/snow: set SPI clock rate in romstage mainDavid Hendricks
This moves the setting of SPI clock rate into romstage's main, which allows us to eliminate a bunch of dependencies from the bootblock (about 7KB worth). Change-Id: I371499bb4af6a6aa838294bc56f9dbc21864957a Signed-off-by: David Hendricks <dhendrix@chromium.org> Reviewed-on: http://review.coreboot.org/2346 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-02-10armv7/exynos5250: place .id between .start and bootblock mainDavid Hendricks
This places the .id section toward the lower region of the coreboot image, before the bootblock. It's easier for humans to find by dumping the image and it also eliminates ID_SECTION_OFFSET which is currently the upper bound on our image size. Change-Id: I7d737b901dac659ddf9aa437cee5dc32f1080546 Signed-off-by: David Hendricks <dhendrix@chromium.org> Reviewed-on: http://review.coreboot.org/2345 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-02-10armv7: make bootblock linker script more explicitDavid Hendricks
This adds a .bl1 and .start symbol that is placed at the beginning of the .rom section. The goal is to move the .id section in between the reset vector and bootblock_main. Change-Id: Ie732ce656d697c059cc0fa40c844b39f53fc214c Signed-off-by: David Hendricks <dhendrix@chromium.org> Reviewed-on: http://review.coreboot.org/2344 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-02-09armv7/exynos: make BL1_SIZE_KB consistent with numbers used...David Hendricks
The Kconfig variable indicates KB, but the number used was bytes. Let's just assume KB is correct for now. Change-Id: I910c126104f0222fc48b70a18df943f2afddeca3 Signed-off-by: David Hendricks <dhendrix@chromium.org> Reviewed-on: http://review.coreboot.org/2341 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-02-09speedstep: Deduplicate some MSR identifiersPatrick Georgi
In particular: MSR_PMG_CST_CONFIG_CONTROL MSR_PMG_IO_BASE_ADDR MSR_PMG_IO_CAPTURE_ADDR Change-Id: Ief2697312f0edf8c45f7d3550a7bedaff1b69dc6 Signed-off-by: Patrick Georgi <patrick@georgi-clan.de> Reviewed-on: http://review.coreboot.org/2337 Tested-by: build bot (Jenkins) Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
2013-02-09romcc: Use default romcc flags for most boardsPatrick Georgi
Except for one board, the flags can be derived from CONFIG_MMX and CONFIG_SSE. Change-Id: I64a11135ee7ce8676f3422b2377069a3fa78e24d Signed-off-by: Patrick Georgi <patrick@georgi-clan.de> Reviewed-on: http://review.coreboot.org/2336 Tested-by: build bot (Jenkins) Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
2013-02-09romcc: Don't use user overridable romcc flags for bootblockPatrick Georgi
The bootblock is typically run before fpu/mmx/sse setup, so we can't rely on -mcpu=p4 and the like to increase the register space. bootblock_romccflags does that for SSE, but they're controlled separately. Change-Id: I2b0609ac18b2394a319bf9bbbee1f77d2e758127 Signed-off-by: Patrick Georgi <patrick@georgi-clan.de> Reviewed-on: http://review.coreboot.org/2335 Tested-by: build bot (Jenkins) Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
2013-02-09armv7: update coreboot tables for armv7David Hendricks
This is a first-pass attempt at cleaning up the coreboot tables for ARM. The most noticable difference is that there is no longer both a high and a low table. Change-Id: I5ba87ad57bf9a697b733511182c0326825071617 Signed-off-by: David Hendricks <dhendrix@chromium.org> Reviewed-on: http://review.coreboot.org/2329 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-02-09snow: do something useful in ramstage()David Hendricks
This cleans up Snow's trivial ramstage, gives it a coreboot table address and calls hardwaremain(). Change-Id: I84c904bcfd57a5f9eb3969de8a496f01e43bc2f6 Signed-off-by: David Hendricks <dhendrix@chromium.org> Reviewed-on: http://review.coreboot.org/2328 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-02-09armv7/snow: add BL1_SIZE_KB and get rid of magic constantsDavid Hendricks
This adds a BL1_SIZE_KB config variable so that we can get rid of some magic constants. Change-Id: I9dbcfb407d3f8e367be5d943e95b032ce88b0ad0 Signed-off-by: David Hendricks <dhendrix@chromium.org> Reviewed-on: http://review.coreboot.org/2332 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-02-09armv7: include $(obj)/config.h when building bootblockDavid Hendricks
Explicitly including it allows us to get rid of some magic constants in the bootblock linker script. Change-Id: I095899babc997addce6b383f00e5ebf135e99d5e Signed-off-by: David Hendricks <dhendrix@chromium.org> Reviewed-on: http://review.coreboot.org/2331 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-02-09DEBUG_CBFS should not depend on TPMDavid Hendricks
This seemed to have been introduced in fe422184. Change-Id: I4f9ecfbec42aa8c0bb8887675a3add8951645b98 Signed-off-by: David Hendricks <dhendrix@chromium.org> Reviewed-on: http://review.coreboot.org/2327 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-02-09document Intel VMX locking behaviorMike Frysinger
Add a comment explaining that the existing lock bit logic is correct and "as designed" even though the manual states otherwise. This way people don't have to "just know" what is going on. Change-Id: I14e6763abfe339e034037b73db01d4ee634bb34d Signed-off-by: Mike Frysinger <vapier@chromium.org> Reviewed-on: http://review.coreboot.org/2326 Tested-by: build bot (Jenkins) Reviewed-by: Peter Stuge <peter@stuge.se>
2013-02-08VBE: Skip graphics mode setting for non-VGA devicesStefan Reinauer
This hit me when running the latest Qemu with coreboot: First the graphics OPROM is running, then an iPXE OPROM. The iPXE OPROM has no int10 support (obviously) so calling vbe_set_graphics() wipes the framebuffer information from the coreboot table. Change-Id: Ie0453c4a908ea4a6216158f663407a3e72ce4d34 Signed-off-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/2325 Reviewed-by: Patrick Georgi <patrick@georgi-clan.de> Tested-by: build bot (Jenkins) Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
2013-02-08console: Fix using CMOS for optionsPatrick Georgi
Just a tiny mistake, but it made the console driver assume that CMOS data isn't available. Change-Id: I4e6f53e9ed59024de7b09333f82f0ce3235ef8f6 Signed-off-by: Patrick Georgi <patrick.georgi@secunet.com> Reviewed-on: http://review.coreboot.org/2323 Reviewed-by: Christian Gmeiner <christian.gmeiner@gmail.com> Tested-by: build bot (Jenkins) Reviewed-by: Peter Stuge <peter@stuge.se>
2013-02-08oprom: fix compilation for Qemu target (and possibly others)Stefan Reinauer
Not sure why this didn't bite us earlier.. src/device/oprom/realmode/x86.c: In function 'fill_lb_framebuffer': src/device/oprom/realmode/x86.c:272:13: error: dereferencing pointer to incomplete type src/device/oprom/realmode/x86.c:274:13: error: dereferencing pointer to incomplete type src/device/oprom/realmode/x86.c:275:13: error: dereferencing pointer to incomplete type src/device/oprom/realmode/x86.c:276:13: error: dereferencing pointer to incomplete type src/device/oprom/realmode/x86.c:278:13: error: dereferencing pointer to incomplete type src/device/oprom/realmode/x86.c:280:13: error: dereferencing pointer to incomplete type src/device/oprom/realmode/x86.c:281:13: error: dereferencing pointer to incomplete type src/device/oprom/realmode/x86.c:283:13: error: dereferencing pointer to incomplete type src/device/oprom/realmode/x86.c:284:13: error: dereferencing pointer to incomplete type src/device/oprom/realmode/x86.c:286:13: error: dereferencing pointer to incomplete type src/device/oprom/realmode/x86.c:287:13: error: dereferencing pointer to incomplete type src/device/oprom/realmode/x86.c:289:13: error: dereferencing pointer to incomplete type src/device/oprom/realmode/x86.c:290:13: error: dereferencing pointer to incomplete type Change-Id: Ie3b0f731a7b995e954a26e745b07fc122088ca9f Signed-off-by: Stefan Reinauer <reinauer@google.com> Reviewed-on: http://review.coreboot.org/2321 Tested-by: build bot (Jenkins) Reviewed-by: David Hendricks <dhendrix@chromium.org>
2013-02-08armv7/snow: Move clock initialization from bootblock to romstage.Hung-Te Lin
Exynos system clock can be initialized before RAM init, not necessary to be in the very beginning (boot block). This helps reducing bootblock dependency. Verified to boot on armv7/snow. Note: this patch was originally introduced in 2308, but there were some ordering issues so it was reverted. Change-Id: Ibc91c0e26ea8881751fc088754f5c6161d011b68 Signed-off-by: Hung-Te Lin <hungte@chromium.org> Signed-off-by: David Hendricks <dhendrix@chromium.org> Reviewed-on: http://review.coreboot.org/2320 Tested-by: build bot (Jenkins) Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
2013-02-08armv7: Use same console initialization procedure for all ARM stagesHung-Te Lin
Use same console initialization procedure for all ARM stages (bootblock, romstage, and ramstage): #include <console/console.h> ... console_init() ... printk(level, format, ...) Verified to boot on armv7/snow with console messages in all stages. Change-Id: Idd689219035e67450ea133838a2ca02f8d74557e Signed-off-by: Hung-Te Lin <hungte@chromium.org> Signed-off-by: David Hendricks <dhendrix@chromium.org> Reviewed-on: http://review.coreboot.org/2301 Tested-by: build bot (Jenkins) Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
2013-02-08console: Only print romstage messages with EARLY_CONSOLE enabled.Hung-Te Lin
Revise console source file dependency (especially for EARLY_CONSOLE) and interpret printk/console_init according to EARLY_CONSOLE setting (no-ops if EARLY_CONSOLE is not defined). Verified to boot on x86/qemu and armv7/snow. Disabling EARLY_CONSOLE correctly stops romstage messages on x86/qemu (armv7/snow needs more changes to work). Change-Id: Idbbd3a26bc1135c9d3ae282aad486961fb60e0ea Signed-off-by: Hung-Te Lin <hungte@chromium.org> Reviewed-on: http://review.coreboot.org/2300 Tested-by: build bot (Jenkins) Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net> Reviewed-by: David Hendricks <dhendrix@chromium.org>
2013-02-08console: Always allow setting "EARLY_CONSOLE" configuration.Hung-Te Lin
Early console should always be allowed to be turned on / off (for generating production and debug versions), and should not be enforced by "select" Kconfig rule. A new "DEFAULT_EARLY_CONSOLE" is introduced for devices to select if they prefer early console output by default. Verified Kconfig value on qemu/x86 (default y by CACHE_AS_RAM), snow/x86 (default y by EXYNOS5 config), and intel/jarrell (default n). Change-Id: Ib1cc76d4ec115a302b95e7317224f1a40d1ab035 Signed-off-by: Hung-Te Lin <hungte@chromium.org> Reviewed-on: http://review.coreboot.org/2307 Tested-by: build bot (Jenkins) Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net> Reviewed-by: David Hendricks <dhendrix@chromium.org>
2013-02-08exynos/snow: Configure UART peripheral during console initialization.Hung-Te Lin
For Exynos platforms, the UART component on pinmux must be first selected and configured. This should be done as part of UART console initialization. Note, that the current implementation hard-codes the device index as UART3, while the base port can be assigned to different device in Kconfig. This will be fixed later. Verified to work on armv7/snow. Change-Id: Ie63e76e2dac09fec1132573d1b0027fce55333a1 Signed-off-by: Hung-Te Lin <hungte@chromium.org> Reviewed-on: http://review.coreboot.org/2315 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com> Reviewed-by: David Hendricks <dhendrix@chromium.org>
2013-02-08console: Revise serial console configuration names.Hung-Te Lin
The console drivers (especially serial drivers) in Kconfig were named in different styles. This change will rename configuration names to a better naming style. - EARLY_CONSOLE: Enable output in pre-ram stage. (Renamed from EARLY_SERIAL_CONSOLE because it also supports non-serial) - CONSOLE_SERIAL: Enable serial output console, from one of the serial drivers. (Renamed from SERIAL_CONSOLE because other non-serial drivers are named as CONSOLE_XXX like CONSOLE_CBMEM) - CONSOLE_SERIAL_UART: Device-specific UART driver. (Renamed from CONSOLE_SERIAL_NONSTANDARD_MEM because it may be not memory-mapped) - HAVE_UART_SPECIAL: A dependency for CONSOLE_SERIAL_UART. Verified to boot on x86/qemu and armv7/snow, and still seeing console messages in romstage for both platforms. Change-Id: I4bea3c8fea05bbb7d78df6bc22f82414ac66f973 Signed-off-by: Hung-Te Lin <hungte@chromium.org> Reviewed-on: http://review.coreboot.org/2299 Tested-by: build bot (Jenkins) Reviewed-by: David Hendricks <dhendrix@chromium.org>
2013-02-08Revert "armv7/snow: Move clock initialization from bootblock to romstage."David Hendricks
This reverts commit 9029f4b63f6d0e29bf1608e666cdb025de45ca24 This patch needs to go at the end of the UART patch set. Sorry 'bout the confusion! Change-Id: I5702c7d6130daf95776f2c15d24e5d253691cefd Reviewed-on: http://review.coreboot.org/2319 Tested-by: build bot (Jenkins) Reviewed-by: David Hendricks <dhendrix@chromium.org>
2013-02-07armv7/snow: Move clock initialization from bootblock to romstage.Hung-Te Lin
Exynos system clock can be initialized before RAM init, not necessary to be in the very beginning (boot block). This helps reducing bootblock dependency. Verified to boot on armv7/snow. Change-Id: Ic863e222871a157ba4279a673775b1e18c6eac0d Signed-off-by: Hung-Te Lin <hungte@chromium.org> Reviewed-on: http://review.coreboot.org/2308 Tested-by: build bot (Jenkins) Reviewed-by: David Hendricks <dhendrix@chromium.org>
2013-02-07YABEL: use system {in,out}[bwl] on x86Patrick Georgi
The prototypes that were recently defined only work for the internal implementations. Change-Id: Ib34bb75a0b882533da550b9cd17cd777c2463e02 Signed-off-by: Patrick Georgi <patrick.georgi@secunet.com> Reviewed-on: http://review.coreboot.org/2318 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-02-07armv7/snow: Remove power_init from bootblock.Hung-Te Lin
The power_init is not required on Exynos 5250 (snow) in bootblock stage. To get a cleaner and faster bootblock, we can remove it. Note, power_init internally calls max77686 and s3c24x0_i2c, so both files are also removed. Verified to boot on armv7/snow. Change-Id: I5b15dfe5ac7bf4650565fea0afefc94a228ece29 Signed-off-by: Hung-Te Lin <hungte@chromium.org> Reviewed-on: http://review.coreboot.org/2317 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-02-07armv7/snow: Remove redundant I2C initialization calls in bootblock.Hung-Te Lin
The I2C initialization (on component MAX77688) is already done in power_init, so we should not need an explicit call inside bootblock. Verified to boot on armv7/snow. Change-Id: I68c248a8b5fee4ab838b2fb708649e112559cc41 Signed-off-by: Hung-Te Lin <hungte@chromium.org> Reviewed-on: http://review.coreboot.org/2316 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-02-07armv7: Clean up arm/snow bootblock build process.Hung-Te Lin
Remove duplicated / testing code and share more driver for bootblock, romstage and ramstage. The __PRE_RAM__ is now also defined in bootblock build stage, since bootblock is executed before RAM is initialized. Change-Id: I4f5469b1545631eee1cf9f2f5df93cbe3a58268b Signed-off-by: Hung-Te Lin <hungte@chromium.org> Reviewed-on: http://review.coreboot.org/2282 Tested-by: build bot (Jenkins) Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
2013-02-07cbfs: Fix CBFS max size calculation.Hung-Te Lin
For x86, the old CBFS search behavior was to bypass bootblock and we should keep that. This will speed up searching if a file does not exist in CBFS. For arm, the size in header is correct now so we can remove the hack by CONFIG_ROM_SIZE. Change-Id: I541961bc4dd083a583f8a80b69e293694fb055ef Signed-off-by: Hung-Te Lin <hungte@chromium.org> Reviewed-on: http://review.coreboot.org/2292 Tested-by: build bot (Jenkins) Reviewed-by: David Hendricks <dhendrix@chromium.org> Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
2013-02-06armv7: Clean up: remove deprecated SPL.Hung-Te Lin
"SPL" from U-Boot is deprecated by bootblock in coreboot/arm, so we don't need it anymore. Change-Id: Id16877075d0b870839a10160073ad70777a2af0a Signed-off-by: Hung-Te Lin <hungte@chromium.org> Reviewed-on: http://review.coreboot.org/2297 Tested-by: build bot (Jenkins) Reviewed-by: David Hendricks <dhendrix@chromium.org> Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
2013-02-06snow: remove dead code from bootblockDavid Hendricks
This attempts to clean out some dead code which was copy + pasted into Snow's bootblock.c file, along with some unnecessary headers. Change-Id: If9f157a52395a047c249a2a6385e0e8ddf310e59 Signed-off-by: David Hendricks <dhendrix@chromium.org> Reviewed-on: http://review.coreboot.org/2291 Tested-by: build bot (Jenkins)
2013-02-06armv7: Clean up: replace hang() by hlt().Hung-Te Lin
hang() is the legacy function from U-boot and should be replaced by hlt() in coreboot. Change-Id: I0f390b1b6f9ff71487ea36cf16c462724b66d8ca Signed-off-by: Hung-Te Lin <hungte@chromium.org> Reviewed-on: http://review.coreboot.org/2298 Tested-by: build bot (Jenkins) Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org> Reviewed-by: David Hendricks <dhendrix@chromium.org>
2013-02-06AMD Fam14 - Fix warningsMartin Roth
Added casts and a couple of #ifdefs to fix the warnings in the vendorcode/amd/agesa/f14 codebase. This will allow us to re-enable 'all warnings being treated as errors' in boards such as Persimmon that are using this code. That change will follow. These are the warnings that are fixed by this patch: src/vendorcode/amd/agesa/f14/Legacy/Proc/hobTransfer.c: In function 'CopyHeapToTempRamAtPost': src/vendorcode/amd/agesa/f14/Legacy/Proc/hobTransfer.c:219:28: warning: cast to pointer from integer of different size [-Wint-to-pointer-cast] src/vendorcode/amd/agesa/f14/Legacy/Proc/hobTransfer.c: In function 'CopyHeapToMainRamAtPost': src/vendorcode/amd/agesa/f14/Legacy/Proc/hobTransfer.c:372:30: warning: comparison between pointer and integer [enabled by default] src/vendorcode/amd/agesa/f14/Legacy/Proc/hobTransfer.c:381:33: warning: cast from pointer to integer of different size [-Wpointer-to-int-cast] src/vendorcode/amd/agesa/f14/Proc/CPU/cpuApicUtilities.c: In function 'ApUtilSetupIdtForHlt': src/vendorcode/amd/agesa/f14/Proc/CPU/cpuApicUtilities.c:863:19: warning: cast from pointer to integer of different size [-Wpointer-to-int-cast] src/vendorcode/amd/agesa/f14/Proc/CPU/cpuApicUtilities.c:872:18: warning: cast from pointer to integer of different size [-Wpointer-to-int-cast] src/vendorcode/amd/agesa/f14/Proc/CPU/cpuMicrocodePatch.c: In function 'LoadMicrocode': src/vendorcode/amd/agesa/f14/Proc/CPU/cpuMicrocodePatch.c:211:28: warning: cast from pointer to integer of different size [-Wpointer-to-int-cast] src/vendorcode/amd/agesa/f14/Proc/CPU/heapManager.c: In function 'HeapManagerInit': src/vendorcode/amd/agesa/f14/Proc/CPU/heapManager.c:167:52: warning: cast from pointer to integer of different size [-Wpointer-to-int-cast] src/vendorcode/amd/agesa/f14/Proc/CPU/heapManager.c:183:14: warning: cast from pointer to integer of different size [-Wpointer-to-int-cast] src/vendorcode/amd/agesa/f14/Proc/CPU/heapManager.c: In function 'HeapGetBaseAddress': src/vendorcode/amd/agesa/f14/Proc/CPU/heapManager.c:669:17: warning: cast to pointer from integer of different size [-Wint-to-pointer-cast] src/vendorcode/amd/agesa/f14/Proc/CPU/heapManager.c:676:19: warning: cast to pointer from integer of different size [-Wint-to-pointer-cast] src/vendorcode/amd/agesa/f14/Proc/CPU/heapManager.c:683:23: warning: cast from pointer to integer of different size [-Wpointer-to-int-cast] src/vendorcode/amd/agesa/f14/Proc/CPU/heapManager.c:684:23: warning: cast to pointer from integer of different size [-Wint-to-pointer-cast] src/vendorcode/amd/agesa/f14/Proc/CPU/heapManager.c:687:23: warning: assignment makes integer from pointer without a cast [enabled by default] src/vendorcode/amd/agesa/f14/Proc/CPU/heapManager.c:691:21: warning: assignment makes integer from pointer without a cast [enabled by default] src/vendorcode/amd/agesa/f14/Proc/CPU/heapManager.c:696:3: warning: return makes pointer from integer without a cast [enabled by default] In file included from src/mainboard/amd/persimmon/agesawrapper.h:30:0, from src/northbridge/amd/agesa/family14/northbridge.c:36: src/vendorcode/amd/agesa/f14/AGESA.h:1132:0: warning: "TOP_MEM" redefined [enabled by default] In file included from src/northbridge/amd/agesa/family14/northbridge.c:34:0: src/include/cpu/amd/mtrr.h:31:0: note: this is the location of the previous definition In file included from src/mainboard/amd/persimmon/agesawrapper.h:30:0, from src/northbridge/amd/agesa/family14/northbridge.c:36: src/vendorcode/amd/agesa/f14/AGESA.h:1133:0: warning: "TOP_MEM2" redefined [enabled by default] In file included from src/northbridge/amd/agesa/family14/northbridge.c:34:0: src/include/cpu/amd/mtrr.h:34:0: note: this is the location of the previous definition Verified on persimmon. Change-Id: I1671b191c72dfc1d63ada41126ae3418bc8f86ae Signed-off-by: Martin Roth <martin.roth@se-eng.com> Reviewed-on: http://review.coreboot.org/2293 Tested-by: build bot (Jenkins) Reviewed-by: Dave Frodin <dave.frodin@se-eng.com> Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net> Reviewed-by: Steven Sherk <steven.sherk@se-eng.com>
2013-02-06bootblock: Reduce register loadPatrick Georgi
The common part of the bootblock resets the nvram data if it's found to be invalid. Since that code is compiled with romcc in i386 mode, there's a shortage on registers. Try to reduce the strain by doing things smarter: cmos_write_inner is the same as cmos_write, just that it doesn't check if the RTC is disabled. Since we just disabled it before, we can assume that it is so. Change-Id: Ic85eb2a5df949d1c1aff654bc1b40d6f2ff71756 Signed-off-by: Patrick Georgi <patrick@georgi-clan.de> Reviewed-on: http://review.coreboot.org/2296 Tested-by: build bot (Jenkins) Reviewed-by: Christian Gmeiner <christian.gmeiner@gmail.com>
2013-02-06armv7: Prevent CBFS data overlapping bootblock.Hung-Te Lin
For arm/snow, current bootblock is larger than previously assigned CBFS offset and will fail to boot. To prevent this happening again in future, cbfstool now checks if CBFS will overlap bootblock. A sample error message: E: Bootblock (0x0+0x71d4) overlap CBFS data (0x5000) E: Failed to create build/coreboot.pre1.tmp. arm/snow offset is also enlarged and moved to Kconfig variable. Change-Id: I4556aef27ff716556040312ae8ccb78078abc82d Signed-off-by: Hung-Te Lin <hungte@chromium.org> Reviewed-on: http://review.coreboot.org/2295 Tested-by: build bot (Jenkins) Reviewed-by: David Hendricks <dhendrix@chromium.org>
2013-02-06cbfs: Revise debug messages.Hung-Te Lin
Some variables are using incorrect data type in debug messages. Also corrects a typo (extra 'x'). Change-Id: Ia3014ea018f8c1e4733c54a7d9ee196d0437cfbb Signed-off-by: Hung-Te Lin <hungte@chromium.org> Reviewed-on: http://review.coreboot.org/2294 Tested-by: build bot (Jenkins) Reviewed-by: David Hendricks <dhendrix@chromium.org>
2013-02-06replace uchar and uint with standard types in generic i2c headerDavid Hendricks
Change-Id: Ie72985bb5291bcef2e837a2f4f2ec929a0c086ce Signed-off-by: David Hendricks <dhendrix@chromium.org> Reviewed-on: http://review.coreboot.org/2290 Tested-by: build bot (Jenkins) Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
2013-02-06exynos/snow: Move core/memory clock-related and board ID codeDavid Hendricks
This patch moves ARM core and DRAM timing functions around to simplify the dependencies for system_clock_init(). The original code was architected such that the system_clock_init() function called other functions to obtain core and memory timings. Due to the way memory timing information must be obtained on Snow, which entails decoding platform-specific board straps, the bottom- up approach resulted in having the low-level clock init code implicitly depend on board and vendor-specific info: main() ->system_clock_init() -> get_arm_ratios() -> CPU-specific code -> clock_get_mem_timings() -> board_get_revision() -> read GPIOs (3-state logic) -> Decode GPIOs in a vendor-specific manner -> Choose memory timings from module-specific look-up table ...then proceed to init clocks ...come back to main() The new approach gathers all board and vendor-specific info in a more appropriate location and passes it into system_clock_init(): main() -> get_arm_ratios() -> CPU-specific code -> get_mem_timings() -> board_get_config() -> read GPIOs (3-state logic) -> Decode GPIOs in a vendor-specific manner -> Choose memory timings from module-specific look-up table -> system_clock_init() ...back to main() Change-Id: Ie237ebff76fc2d8a4d2f4577a226ac3909e4d4e8 Signed-off-by: David Hendricks <dhendrix@chromium.org> Reviewed-on: http://review.coreboot.org/2271 Tested-by: build bot (Jenkins) Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
generated by cgit v1.2.3 (git 2.25.1) at 2024-11-02 14:36:11 +0000