August 2013 - OpenBIOS - mail.coreboot.org

[commit] r1209 - trunk/openbios-devel/arch/sparc32
by repository service Dec. 8, 2013

Dec. 8, 2013

Author: mcayland Date: Mon Aug 19 09:40:12 2013 New Revision: 1209 URL: http://tracker.coreboot.org/trac/openbios/changeset/1209 Log: SPARC32: fix romvec stdin/stdout field initialisation Previously the romvec stdin/stdout ihandles were configured so that they were set to the current stdin and stdout paths at initialisation time. Unfortunately as stdin/stdout can be changed with the input and output words, they can end up pointing to invalid ihandles causing a crash when trying to output to the console. Fix this by pointing the romvec structure to the address of the stdin/stdout variables so that they are always in sync. Similarly we also resolve the text path stdin/stdout variables at boot time to ensure that they will never be stale. Signed-off-by: Mark Cave-Ayland <mark.cave-ayland(a)ilande.co.uk> Modified: trunk/openbios-devel/arch/sparc32/boot.c trunk/openbios-devel/arch/sparc32/romvec.c Modified: trunk/openbios-devel/arch/sparc32/boot.c ============================================================================== --- trunk/openbios-devel/arch/sparc32/boot.c Mon Aug 19 09:40:10 2013 (r1208) +++ trunk/openbios-devel/arch/sparc32/boot.c Mon Aug 19 09:40:12 2013 (r1209) @@ -23,7 +23,7 @@ void go(void) { ucell address, type, size; - int image_retval = 0, proplen, target, device; + int image_retval = 0, intprop, proplen, target, device; phandle_t chosen; char *prop, *id, *name; static char bootpathbuf[128], bootargsbuf[128], buf[128]; @@ -40,8 +40,19 @@ needs to be set up to pass certain parameters using a C struct. Hence this section extracts the relevant boot information and places it in obp_arg. */ - /* Get the name of the selected boot device, along with the device and unit number */ + /* Get the stdin and stdout paths */ chosen = find_dev("/chosen"); + intprop = get_int_property(chosen, "stdin", &proplen); + PUSH(intprop); + fword("get-instance-path"); + ((struct linux_romvec *)romvec)->pv_stdin = pop_fstr_copy(); + + intprop = get_int_property(chosen, "stdout", &proplen); + PUSH(intprop); + fword("get-instance-path"); + ((struct linux_romvec *)romvec)->pv_stdout = pop_fstr_copy(); + + /* Get the name of the selected boot device, along with the device and unit number */ prop = get_property(chosen, "bootpath", &proplen); strncpy(bootpathbuf, prop, proplen); prop = get_property(chosen, "bootargs", &proplen); Modified: trunk/openbios-devel/arch/sparc32/romvec.c ============================================================================== --- trunk/openbios-devel/arch/sparc32/romvec.c Mon Aug 19 09:40:10 2013 (r1208) +++ trunk/openbios-devel/arch/sparc32/romvec.c Mon Aug 19 09:40:12 2013 (r1209) @@ -25,7 +25,6 @@ #endif char obp_stdin, obp_stdout; -static int obp_fd_stdin, obp_fd_stdout; const char *obp_stdin_path, *obp_stdout_path; struct linux_arguments_v0 obp_arg; @@ -498,15 +497,12 @@ romvec0.pv_v2bootargs.bootpath = &bootpath; romvec0.pv_v2bootargs.bootargs = &obp_arg.argv[1]; - romvec0.pv_v2bootargs.fd_stdin = &obp_fd_stdin; - romvec0.pv_v2bootargs.fd_stdout = &obp_fd_stdout; - push_str(obp_stdin_path); - fword("open-dev"); - obp_fd_stdin = POP(); - push_str(obp_stdout_path); - fword("open-dev"); - obp_fd_stdout = POP(); + /* Point fd_stdin/fd_stdout to the Forth stdin/stdout variables */ + fword("stdin"); + romvec0.pv_v2bootargs.fd_stdin = cell2pointer(POP()); + fword("stdout"); + romvec0.pv_v2bootargs.fd_stdout = cell2pointer(POP()); romvec0.v3_memalloc = obp_memalloc_handler;

3 6

[PATCH 1/2] PPC: Fix PCI configuration space address access
by Mark Cave-Ayland Sept. 2, 2013

Sept. 2, 2013

This patch corrects two long-standing bugs with PPC PCI configuration space access. Firstly fix the calculation of PCI configuration space addresses by the PCI_ADDR macro; this was incorrectly using arch->cfg_base which is the mapped address and has nothing to do with the PCI configuration space address. Instead just set bit 31 to initiate a configuration cycle as per the PCI specification. Secondly, fix pci_config_read32() and pci_config_write16() which were incorrectly adding the register offset to the PCI IO dataport address causing them to write into unknown address space for registers > 0. It appears that this only worked purely by coincidence with QEMU due to the way in which the configuration address was calculated for an oversized PCI configuration IO dataport MemoryRegion. Reported-by: Hervé Poussineau <hpoussin(a)reactos.org> CC: Hervé Poussineau <hpoussin(a)reactos.org> CC: Andreas Färber <afaerber(a)suse.de> Signed-off-by: Mark Cave-Ayland <mark.cave-ayland(a)ilande.co.uk> --- openbios-devel/include/arch/ppc/pci.h | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/openbios-devel/include/arch/ppc/pci.h b/openbios-devel/include/arch/ppc/pci.h index cccefb1..d96bd7e 100644 --- a/openbios-devel/include/arch/ppc/pci.h +++ b/openbios-devel/include/arch/ppc/pci.h @@ -12,7 +12,7 @@ /* PCI Configuration Mechanism #1 */ #define PCI_ADDR(bus, dev, fn) \ - ((pci_addr) (arch->cfg_base \ + ((pci_addr) (0x80000000u \ | (uint32_t) (bus) << 16 \ | (uint32_t) (dev) << 11 \ | (uint32_t) (fn) << 8)) @@ -41,7 +41,7 @@ static inline uint32_t pci_config_read32(pci_addr dev, uint8_t reg) { uint32_t res; out_le32((unsigned *)arch->cfg_addr, dev | reg); - res = in_le32((unsigned *)(arch->cfg_data + reg)); + res = in_le32((unsigned *)(arch->cfg_data)); return res; } @@ -60,7 +60,7 @@ static inline void pci_config_write16(pci_addr dev, uint8_t reg, uint16_t val) static inline void pci_config_write32(pci_addr dev, uint8_t reg, uint32_t val) { out_le32((unsigned *)arch->cfg_addr, dev | reg); - out_le32((unsigned *)(arch->cfg_data + reg), val); + out_le32((unsigned *)(arch->cfg_data), val); } #else /* !PCI_CONFIG_1 */ #error PCI Configuration Mechanism is not specified or implemented -- 1.7.10.4

2 4

Re: [OpenBIOS] [PATCH] ppc: Implement NVRAM support for PReP
by Mark Cave-Ayland Aug. 30, 2013

Aug. 30, 2013

On 27/08/13 11:10, Artyom Tarasenko wrote: > Hi Mark, > > On Mon, Aug 26, 2013 at 11:05 PM, Mark Cave-Ayland > <mark.cave-ayland(a)ilande.co.uk> wrote: >> On 24/08/13 23:06, Andreas Färber wrote: >> >>> Signed-off-by: Andreas Färber<andreas.faerber(a)web.de> >>> --- >>> arch/ppc/qemu/qemu.c | 18 ++++++++++-- >>> config/examples/ppc64_config.xml | 1 + >>> config/examples/ppc_config.xml | 1 + >>> drivers/build.xml | 1 + >>> drivers/m48t59.c | 60 >>> ++++++++++++++++++++++++++++++++++++++++ >>> include/drivers/drivers.h | 3 ++ >>> 6 files changed, 81 insertions(+), 3 deletions(-) >>> create mode 100644 drivers/m48t59.c >>> >>> diff --git a/arch/ppc/qemu/qemu.c b/arch/ppc/qemu/qemu.c >>> index 381affb..1e3994c 100644 >>> --- a/arch/ppc/qemu/qemu.c >>> +++ b/arch/ppc/qemu/qemu.c >>> @@ -80,9 +80,13 @@ printk( const char *fmt, ... ) >>> int arch_nvram_size(void) >>> { >>> if (is_apple()) { >>> +#ifdef CONFIG_DRIVER_MACIO >>> return macio_get_nvram_size(); >>> +#endif >>> } else { >>> - // not implemented >>> +#ifdef CONFIG_DRIVER_M48T59 >>> + return m48t59_nvram_size(); >>> +#endif >>> } >>> return 0; >>> } >>> @@ -90,17 +94,25 @@ int arch_nvram_size(void) >>> void arch_nvram_put(char *buf) >>> { >>> if (is_apple()) { >>> +#ifdef CONFIG_DRIVER_MACIO >>> macio_nvram_put(buf); >>> +#endif >>> } else { >>> - // not implemented >>> +#ifdef CONFIG_DRIVER_M48T59 >>> + m48t59_nvram_put(buf); >>> +#endif >>> } >>> } >>> >>> void arch_nvram_get(char *buf) >>> { >>> if (is_apple()) { >>> +#ifdef CONFIG_DRIVER_MACIO >>> macio_nvram_get(buf); >>> +#endif >>> } else { >>> - // not implemented >>> +#ifdef CONFIG_DRIVER_M48T59 >>> + m48t59_nvram_get(buf); >>> +#endif >>> } >>> } >>> diff --git a/config/examples/ppc64_config.xml >>> b/config/examples/ppc64_config.xml >>> index 5f79c21..04343b4 100644 >>> --- a/config/examples/ppc64_config.xml >>> +++ b/config/examples/ppc64_config.xml >>> @@ -75,3 +75,4 @@ >>> <option name="CONFIG_DRIVER_ESCC" type="boolean" value="true"/> >>> <option name="CONFIG_DRIVER_FW_CFG" type="boolean" value="true"/> >>> <option name="CONFIG_FW_CFG_ADDR" type="integer" value="0xf0000510"/> >>> +<option name="CONFIG_DRIVER_M48T59" type="boolean" value="false"/> >>> diff --git a/config/examples/ppc_config.xml >>> b/config/examples/ppc_config.xml >>> index 5bb789f..2229f00 100644 >>> --- a/config/examples/ppc_config.xml >>> +++ b/config/examples/ppc_config.xml >>> @@ -78,3 +78,4 @@ >>> <option name="CONFIG_DRIVER_ESCC" type="boolean" value="true"/> >>> <option name="CONFIG_DRIVER_FW_CFG" type="boolean" value="true"/> >>> <option name="CONFIG_FW_CFG_ADDR" type="integer" value="0xf0000510"/> >>> +<option name="CONFIG_DRIVER_M48T59" type="boolean" value="true"/> >>> diff --git a/drivers/build.xml b/drivers/build.xml >>> index edec6b5..bcb21ae 100644 >>> --- a/drivers/build.xml >>> +++ b/drivers/build.xml >>> @@ -22,6 +22,7 @@ >>> <object source="pc_serial.c" condition="DRIVER_PC_SERIAL"/> >>> <object source="escc.c" condition="DRIVER_ESCC"/> >>> <object source="fw_cfg.c" condition="DRIVER_FW_CFG"/> >>> +<object source="m48t59.c" condition="DRIVER_M48T59"/> >>> </library> >>> >>> <dictionary name="openbios" target="forth"> >>> diff --git a/drivers/m48t59.c b/drivers/m48t59.c >>> new file mode 100644 >>> index 0000000..cd9e9f6 >>> --- /dev/null >>> +++ b/drivers/m48t59.c >>> @@ -0,0 +1,60 @@ >>> +/* >>> + * OpenBIOS PReP m48t59 driver >>> + * >>> + * Copyright (c) 2013 Andreas Färber >>> + */ >>> + >>> +#include "config.h" >>> + >>> +#include "drivers/drivers.h" >>> +#include "asm/io.h" >>> + >>> +#define PREP_NVRAM_SIZE 0x2000 >>> +#define PREP_NVRAM_ADDR_LO 0x74 >>> +#define PREP_NVRAM_ADDR_HI 0x75 >>> +#define PREP_NVRAM_DATA 0x77 >>> + >>> +static uint8_t m48t59_nvram_read_byte(uint16_t offset) >>> +{ >>> + outb(offset& 0xff, PREP_NVRAM_ADDR_LO); >>> >>> + outb(offset>> 8, PREP_NVRAM_ADDR_HI); >>> + return inb(PREP_NVRAM_DATA); >>> +} >>> + >>> +static void m48t59_nvram_read(uint16_t offset, char *buf, unsigned int >>> len) >>> +{ >>> + unsigned int i; >>> + for (i = 0; i< len; i++) { >>> + buf[i] = m48t59_nvram_read_byte(offset); >>> + } >>> +} >>> + >>> +static void m48t59_nvram_write_byte(uint16_t offset, uint8_t value) >>> +{ >>> + outb(offset& 0xff, PREP_NVRAM_ADDR_LO); >>> >>> + outb(offset>> 8, PREP_NVRAM_ADDR_HI); >>> + outb(value, PREP_NVRAM_DATA); >>> +} >>> + >>> +static void m48t59_nvram_write(uint16_t offset, const char *buf, unsigned >>> int len) >>> +{ >>> + unsigned int i; >>> + for (i = 0; i< len; i++) { >>> + m48t59_nvram_write_byte(offset, buf[i]); >>> + } >>> +} >>> + >>> +int m48t59_nvram_size(void) >>> +{ >>> + return PREP_NVRAM_SIZE; >>> +} >>> + >>> +void m48t59_nvram_put(char *buf) >>> +{ >>> + m48t59_nvram_write(0, buf, PREP_NVRAM_SIZE); >>> +} >>> + >>> +void m48t59_nvram_get(char *buf) >>> +{ >>> + m48t59_nvram_read(0, buf, PREP_NVRAM_SIZE); >>> +} >>> diff --git a/include/drivers/drivers.h b/include/drivers/drivers.h >>> index 47d730f..73da6a2 100644 >>> --- a/include/drivers/drivers.h >>> +++ b/include/drivers/drivers.h >>> @@ -125,6 +125,9 @@ unsigned char keyboard_readdata(void); >>> int macio_get_nvram_size(void); >>> void macio_nvram_put(char *buf); >>> void macio_nvram_get(char *buf); >>> +int m48t59_nvram_size(void); >>> +void m48t59_nvram_put(char *buf); >>> +void m48t59_nvram_get(char *buf); >>> >>> /* drivers/timer.c */ >>> void setup_timers(void); >> >> >> The basic patch looks good, although I have a feeling that the m48t59 is the >> same NVRAM chip used by SPARC64 so there could be some changes required to >> support both. Artyom, any thoughts? > > The m48t59 wired differently on sun4u machines. PReP seems to use PIO, > whereas sun4u uses MMIO for accessing the chip. (It is possible that > both modes are supported on sun4u, but only MMIO is exposed to user > via OBP, but I couldn't find any documentation which would prove or > deny that). > > No objections to this patch from my side. Thanks Artyom. Given that m48t59 could potentially be used across different architectures, on further inspection I'd prefer something a bit more generic and a bit less PReP-specific. As a compromise, how about renaming the PREP_NVRAM_* constants to M48T59_NVRAM_* and moving them to a .h file somewhere under include/arch/ppc? Then at least if someone tries to re-use the driver in a different architecture, it will fail to compile which will serve as a reminder that some alterations to the interface will be required. ATB, Mark.

2 1

[PATCH] ppc: Implement NVRAM support for PReP
by Andreas Färber Aug. 27, 2013

Aug. 27, 2013

Signed-off-by: Andreas Färber <andreas.faerber(a)web.de> --- arch/ppc/qemu/qemu.c | 18 ++++++++++-- config/examples/ppc64_config.xml | 1 + config/examples/ppc_config.xml | 1 + drivers/build.xml | 1 + drivers/m48t59.c | 60 ++++++++++++++++++++++++++++++++++++++++ include/drivers/drivers.h | 3 ++ 6 files changed, 81 insertions(+), 3 deletions(-) create mode 100644 drivers/m48t59.c diff --git a/arch/ppc/qemu/qemu.c b/arch/ppc/qemu/qemu.c index 381affb..1e3994c 100644 --- a/arch/ppc/qemu/qemu.c +++ b/arch/ppc/qemu/qemu.c @@ -80,9 +80,13 @@ printk( const char *fmt, ... ) int arch_nvram_size(void) { if (is_apple()) { +#ifdef CONFIG_DRIVER_MACIO return macio_get_nvram_size(); +#endif } else { - // not implemented +#ifdef CONFIG_DRIVER_M48T59 + return m48t59_nvram_size(); +#endif } return 0; } @@ -90,17 +94,25 @@ int arch_nvram_size(void) void arch_nvram_put(char *buf) { if (is_apple()) { +#ifdef CONFIG_DRIVER_MACIO macio_nvram_put(buf); +#endif } else { - // not implemented +#ifdef CONFIG_DRIVER_M48T59 + m48t59_nvram_put(buf); +#endif } } void arch_nvram_get(char *buf) { if (is_apple()) { +#ifdef CONFIG_DRIVER_MACIO macio_nvram_get(buf); +#endif } else { - // not implemented +#ifdef CONFIG_DRIVER_M48T59 + m48t59_nvram_get(buf); +#endif } } diff --git a/config/examples/ppc64_config.xml b/config/examples/ppc64_config.xml index 5f79c21..04343b4 100644 --- a/config/examples/ppc64_config.xml +++ b/config/examples/ppc64_config.xml @@ -75,3 +75,4 @@ <option name="CONFIG_DRIVER_ESCC" type="boolean" value="true"/> <option name="CONFIG_DRIVER_FW_CFG" type="boolean" value="true"/> <option name="CONFIG_FW_CFG_ADDR" type="integer" value="0xf0000510"/> + <option name="CONFIG_DRIVER_M48T59" type="boolean" value="false"/> diff --git a/config/examples/ppc_config.xml b/config/examples/ppc_config.xml index 5bb789f..2229f00 100644 --- a/config/examples/ppc_config.xml +++ b/config/examples/ppc_config.xml @@ -78,3 +78,4 @@ <option name="CONFIG_DRIVER_ESCC" type="boolean" value="true"/> <option name="CONFIG_DRIVER_FW_CFG" type="boolean" value="true"/> <option name="CONFIG_FW_CFG_ADDR" type="integer" value="0xf0000510"/> + <option name="CONFIG_DRIVER_M48T59" type="boolean" value="true"/> diff --git a/drivers/build.xml b/drivers/build.xml index edec6b5..bcb21ae 100644 --- a/drivers/build.xml +++ b/drivers/build.xml @@ -22,6 +22,7 @@ <object source="pc_serial.c" condition="DRIVER_PC_SERIAL"/> <object source="escc.c" condition="DRIVER_ESCC"/> <object source="fw_cfg.c" condition="DRIVER_FW_CFG"/> + <object source="m48t59.c" condition="DRIVER_M48T59"/> </library> <dictionary name="openbios" target="forth"> diff --git a/drivers/m48t59.c b/drivers/m48t59.c new file mode 100644 index 0000000..cd9e9f6 --- /dev/null +++ b/drivers/m48t59.c @@ -0,0 +1,60 @@ +/* + * OpenBIOS PReP m48t59 driver + * + * Copyright (c) 2013 Andreas Färber + */ + +#include "config.h" + +#include "drivers/drivers.h" +#include "asm/io.h" + +#define PREP_NVRAM_SIZE 0x2000 +#define PREP_NVRAM_ADDR_LO 0x74 +#define PREP_NVRAM_ADDR_HI 0x75 +#define PREP_NVRAM_DATA 0x77 + +static uint8_t m48t59_nvram_read_byte(uint16_t offset) +{ + outb(offset & 0xff, PREP_NVRAM_ADDR_LO); + outb(offset >> 8, PREP_NVRAM_ADDR_HI); + return inb(PREP_NVRAM_DATA); +} + +static void m48t59_nvram_read(uint16_t offset, char *buf, unsigned int len) +{ + unsigned int i; + for (i = 0; i < len; i++) { + buf[i] = m48t59_nvram_read_byte(offset); + } +} + +static void m48t59_nvram_write_byte(uint16_t offset, uint8_t value) +{ + outb(offset & 0xff, PREP_NVRAM_ADDR_LO); + outb(offset >> 8, PREP_NVRAM_ADDR_HI); + outb(value, PREP_NVRAM_DATA); +} + +static void m48t59_nvram_write(uint16_t offset, const char *buf, unsigned int len) +{ + unsigned int i; + for (i = 0; i < len; i++) { + m48t59_nvram_write_byte(offset, buf[i]); + } +} + +int m48t59_nvram_size(void) +{ + return PREP_NVRAM_SIZE; +} + +void m48t59_nvram_put(char *buf) +{ + m48t59_nvram_write(0, buf, PREP_NVRAM_SIZE); +} + +void m48t59_nvram_get(char *buf) +{ + m48t59_nvram_read(0, buf, PREP_NVRAM_SIZE); +} diff --git a/include/drivers/drivers.h b/include/drivers/drivers.h index 47d730f..73da6a2 100644 --- a/include/drivers/drivers.h +++ b/include/drivers/drivers.h @@ -125,6 +125,9 @@ unsigned char keyboard_readdata(void); int macio_get_nvram_size(void); void macio_nvram_put(char *buf); void macio_nvram_get(char *buf); +int m48t59_nvram_size(void); +void m48t59_nvram_put(char *buf); +void m48t59_nvram_get(char *buf); /* drivers/timer.c */ void setup_timers(void); -- 1.8.1.4

3 2

[commit] r1224 - in trunk/openbios-devel: arch/ppc/qemu drivers include/drivers
by repository service Aug. 24, 2013

Aug. 24, 2013

Author: afaerber Date: Sat Aug 24 22:00:36 2013 New Revision: 1224 URL: http://tracker.coreboot.org/trac/openbios/changeset/1224 Log: macio: Rename arch_nvram_{put,get,size}() PReP does not have MacIO, so don't call its functions arch_*. Instead, dispatch from arch_* functions to macio_* functions via is_apple(). Signed-off-by: Andreas Färber <andreas.faerber(a)web.de> Modified: trunk/openbios-devel/arch/ppc/qemu/qemu.c trunk/openbios-devel/drivers/macio.c trunk/openbios-devel/include/drivers/drivers.h Modified: trunk/openbios-devel/arch/ppc/qemu/qemu.c ============================================================================== --- trunk/openbios-devel/arch/ppc/qemu/qemu.c Mon Aug 19 14:58:12 2013 (r1223) +++ trunk/openbios-devel/arch/ppc/qemu/qemu.c Sat Aug 24 22:00:36 2013 (r1224) @@ -76,3 +76,31 @@ } return i; } + +int arch_nvram_size(void) +{ + if (is_apple()) { + return macio_get_nvram_size(); + } else { + // not implemented + } + return 0; +} + +void arch_nvram_put(char *buf) +{ + if (is_apple()) { + macio_nvram_put(buf); + } else { + // not implemented + } +} + +void arch_nvram_get(char *buf) +{ + if (is_apple()) { + macio_nvram_get(buf); + } else { + // not implemented + } +} Modified: trunk/openbios-devel/drivers/macio.c ============================================================================== --- trunk/openbios-devel/drivers/macio.c Mon Aug 19 14:58:12 2013 (r1223) +++ trunk/openbios-devel/drivers/macio.c Sat Aug 24 22:00:36 2013 (r1224) @@ -35,7 +35,7 @@ static char *nvram; int -arch_nvram_size( void ) +macio_get_nvram_size(void) { if (is_oldworld()) return OW_IO_NVRAM_SIZE >> OW_IO_NVRAM_SHIFT; @@ -120,7 +120,7 @@ void -arch_nvram_put( char *buf ) +macio_nvram_put(char *buf) { int i; unsigned int it_shift; @@ -139,7 +139,7 @@ } void -arch_nvram_get( char *buf ) +macio_nvram_get(char *buf) { int i; unsigned int it_shift; Modified: trunk/openbios-devel/include/drivers/drivers.h ============================================================================== --- trunk/openbios-devel/include/drivers/drivers.h Mon Aug 19 14:58:12 2013 (r1223) +++ trunk/openbios-devel/include/drivers/drivers.h Sat Aug 24 22:00:36 2013 (r1224) @@ -122,6 +122,9 @@ unsigned char keyboard_readdata(void); #endif #endif +int macio_get_nvram_size(void); +void macio_nvram_put(char *buf); +void macio_nvram_get(char *buf); /* drivers/timer.c */ void setup_timers(void);

1 0

[PATCH 0/2] CG3 Framebuffer preview
by Mark Cave-Ayland Aug. 19, 2013

Aug. 19, 2013

Finally we have a working CG3 framebuffer with QEMU on SPARC32! This patchset, minus integration of Bob's cpeek patches, should be fairly close to the final implementation. It includes FCode source for a CG3 ROM which can be probed using a patched version of QEMU, and should be applied on top of my previous v2 display patchset. As a teaser, I've created the following images from my testing: http://www.ilande.co.uk/tmp/sol8-1.png http://www.ilande.co.uk/tmp/sol8-2.png http://www.ilande.co.uk/tmp/sol8-3.png http://www.ilande.co.uk/tmp/debian-woody.png Some discussion points: what should be the fallback behaviour if we fail an FCode ROM probe for the display? At the moment I attempt to execute an internal copy of the FCode payload directly, so if people are running OpenBIOS on systems which don't emulate SBus then there should be no difference in behaviour. The only minor issue at the moment is that OpenBIOS will crash if you try to drive the display at 1152x900 rather than 1024x768. This is simply because there isn't enough room in the 2M address space for the extra 200K or so required for the extra screen area. We could probably fix this by trimming out some of the extras such as grubfs (and possibly pre-aligning parts of the memory image) but any ideas gratefully received. If you have access to proprietary ROM images then those should work with the patched QEMU too - please test and let me know. Finally I've pushed my modifed QEMU source with updated OpenBIOS images to github for testing. Simply pass -vga cg3 on the command line and you will be switched onto the new CG3 framebuffer. https://github.com/mcayland/qemu/tree/cg3-preview As always thanks to everyone else for helping to make this possible - enjoy! Mark Cave-Ayland (2): cg3: add new FCode ROM for QEMU's Sun CG3 framebuffer sbus: add probe for display device openbios-devel/arch/sparc32/build.xml | 1 + openbios-devel/drivers/build.xml | 1 + openbios-devel/drivers/cgthree.fs | 154 +++++++++++++++++++++++++++++++++ openbios-devel/drivers/sbus.c | 30 ++++++- openbios-devel/drivers/sbus.fs | 27 ++++-- 5 files changed, 202 insertions(+), 11 deletions(-) create mode 100644 openbios-devel/drivers/cgthree.fs -- 1.7.10.4

1 3

[PATCH 0/3] peek/poke fault handling
by Bob Breuer Aug. 19, 2013

Aug. 19, 2013

Implement data fault handling for cpeek and friends on SPARC32. Keeps the existing behavior for other architectures. This series incorporates contributions and naming suggestions from Mark Cave-Ayland. Bob Breuer (3): Add peek/poke hooks SPARC32: implement data fault handler SPARC32: handle data faults for peek/poke openbios-devel/arch/sparc32/init.fs | 30 ++++++++++++++++++++++++++++++ openbios-devel/arch/sparc32/lib.c | 21 +++++++++++++++++++++ openbios-devel/arch/sparc32/vectors.S | 32 ++++++++++++++++++++++++++++++-- openbios-devel/forth/device/other.fs | 22 ++++++++++++++++------ 4 files changed, 97 insertions(+), 8 deletions(-) -- 1.7.10.4

2 4

[commit] r1223 - in trunk/openbios-devel: drivers forth/admin
by repository service Aug. 19, 2013

Aug. 19, 2013

Author: mcayland Date: Mon Aug 19 14:58:12 2013 New Revision: 1223 URL: http://tracker.coreboot.org/trac/openbios/changeset/1223 Log: sbus: add SBus slot probe functionality Change the SBus logic so that we first probe each slot and attempt to execute any FCode if present. Only if no FCode is found do we fall back to the internal (hardwired) device node generation. Signed-off-by: Mark Cave-Ayland <mark.cave-ayland(a)ilande.co.uk> Modified: trunk/openbios-devel/drivers/sbus.c trunk/openbios-devel/drivers/sbus.fs trunk/openbios-devel/forth/admin/devices.fs Modified: trunk/openbios-devel/drivers/sbus.c ============================================================================== --- trunk/openbios-devel/drivers/sbus.c Mon Aug 19 14:58:09 2013 (r1222) +++ trunk/openbios-devel/drivers/sbus.c Mon Aug 19 14:58:12 2013 (r1223) @@ -141,8 +141,26 @@ static void ob_tcx_init(unsigned int slot, const char *path) { - push_str(path); - fword("find-device"); + char buf[6]; + + printk("No display device located during SBus probe - falling back to internal TCX driver\n"); + + /* Make the sbus node the current instance and active package for probing */ + feval("active-package my-self"); + push_str("/iommu/sbus"); + feval("2dup find-device open-dev to my-self"); + + fword("new-device"); + PUSH(0); + PUSH(0); + snprintf(buf, 6, "%d,0", slot); + push_str(buf); + fword("set-args"); + feval("['] tcx-driver-fcode 2 cells + 1 byte-load"); + fword("finish-device"); + + /* Restore */ + feval("to my-self active-package!"); } static void @@ -241,6 +259,8 @@ as the current instance during probe. */ char buf[6]; + printk("Probing SBus slot %d offset %ld\n", slot, offset); + /* Make the sbus node the current instance and active package for probing */ feval("active-package my-self"); push_str("/iommu/sbus"); @@ -257,13 +277,28 @@ feval("to my-self active-package!"); } +static int +sbus_probe_sucess(void) +{ + /* Return true if the last sbus_probe_self() resulted in + the successful detection and execution of FCode */ + fword("probe-fcode?"); + return POP(); +} + static void sbus_probe_slot_ss5(unsigned int slot, uint64_t base) { - // OpenBIOS and QEMU don't know how to do Sbus probing + /* Probe the slot */ + sbus_probe_self(slot, 0); + + /* If the device was successfully created by FCode then do nothing */ + if (sbus_probe_sucess()) { + return; + } + switch(slot) { case 3: // SUNW,tcx - sbus_probe_self(slot, 0); ob_tcx_init(slot, "/iommu/sbus/SUNW,tcx"); break; case 4: @@ -283,10 +318,16 @@ static void sbus_probe_slot_ss10(unsigned int slot, uint64_t base) { - // OpenBIOS and QEMU don't know how to do Sbus probing + /* Probe the slot */ + sbus_probe_self(slot, 0); + + /* If the device was successfully created by FCode then do nothing */ + if (sbus_probe_sucess()) { + return; + } + switch(slot) { case 2: // SUNW,tcx - sbus_probe_self(slot, 0); ob_tcx_init(slot, "/iommu/sbus/SUNW,tcx"); break; case 0xf: // le, esp, bpp, power-management @@ -302,10 +343,16 @@ static void sbus_probe_slot_ss600mp(unsigned int slot, uint64_t base) { - // OpenBIOS and QEMU don't know how to do Sbus probing + /* Probe the slot */ + sbus_probe_self(slot, 0); + + /* If the device was successfully created by FCode then do nothing */ + if (sbus_probe_sucess()) { + return; + } + switch(slot) { case 2: // SUNW,tcx - sbus_probe_self(slot, 0); ob_tcx_init(slot, "/iommu/sbus/SUNW,tcx"); break; case 0xf: // le, esp, bpp, power-management Modified: trunk/openbios-devel/drivers/sbus.fs ============================================================================== --- trunk/openbios-devel/drivers/sbus.fs Mon Aug 19 14:58:09 2013 (r1222) +++ trunk/openbios-devel/drivers/sbus.fs Mon Aug 19 14:58:12 2013 (r1223) @@ -55,7 +55,7 @@ r> drop ; -: map-out-sbus ( virt ) +: map-out-sbus ( virt size ) " map-out" $call-parent ; @@ -64,15 +64,31 @@ \ ------------------------------------------------------------------------- : probe-self-sbus ( arg-adr arg-len reg-adr reg-len fcode-adr fcode-len -- ) - 2drop - new-device - set-args - - \ Note: this is currently hardcoded to TCX for testing as we don't have - \ cpeek (yet). Once cpeek is in place, adapting this to probe any slot - \ will be faily easy. - " tcx-driver-fcode" $find drop 2 cells + - 1 byte-load - finish-device + 0 to probe-fcode? + + ['] decode-unit-sbus catch if + 2drop 2drop 2drop 2drop + exit + then + + h# 10000 map-in-sbus + + dup cpeek if + dup h# f1 = swap h# fd = or if + new-device + >r set-args r> + dup 1 byte-load + finish-device + + -1 to probe-fcode? + else + nip nip nip nip + ." Invalid FCode start byte" cr + then + else + nip nip nip nip + then + + h# 10000 map-out-sbus ; Modified: trunk/openbios-devel/forth/admin/devices.fs ============================================================================== --- trunk/openbios-devel/forth/admin/devices.fs Mon Aug 19 14:58:09 2013 (r1222) +++ trunk/openbios-devel/forth/admin/devices.fs Mon Aug 19 14:58:12 2013 (r1223) @@ -508,5 +508,8 @@ \ 7.4.11.3 Device probing +\ Set to true if the last probe-self was successful +0 value probe-fcode? + : probe-all ( -- ) ;

1 0

[commit] r1222 - in trunk/openbios-devel: arch/sparc32 drivers
by repository service Aug. 19, 2013

Aug. 19, 2013

Author: mcayland Date: Mon Aug 19 14:58:09 2013 New Revision: 1222 URL: http://tracker.coreboot.org/trac/openbios/changeset/1222 Log: cg3: add new FCode ROM for Sun CG3 framebuffer Signed-off-by: Mark Cave-Ayland <mark.cave-ayland(a)ilande.co.uk> Added: trunk/openbios-devel/drivers/cgthree.fs Modified: trunk/openbios-devel/arch/sparc32/build.xml trunk/openbios-devel/drivers/build.xml Modified: trunk/openbios-devel/arch/sparc32/build.xml ============================================================================== --- trunk/openbios-devel/arch/sparc32/build.xml Mon Aug 19 14:58:06 2013 (r1221) +++ trunk/openbios-devel/arch/sparc32/build.xml Mon Aug 19 14:58:09 2013 (r1222) @@ -4,6 +4,7 @@ <object source="tree.fs" target="forth"/> <object source="init.fs" target="forth"/> <object source="QEMU,tcx.bin" target="fcode" condition="DRIVER_SBUS"/> + <object source="QEMU,cgthree.bin" target="fcode" condition="DRIVER_SBUS"/> </dictionary> <library name="sparc32" type="static" target="target"> Modified: trunk/openbios-devel/drivers/build.xml ============================================================================== --- trunk/openbios-devel/drivers/build.xml Mon Aug 19 14:58:06 2013 (r1221) +++ trunk/openbios-devel/drivers/build.xml Mon Aug 19 14:58:09 2013 (r1222) @@ -31,6 +31,7 @@ </dictionary> <fcode source="tcx.fs" name="QEMU,tcx.bin" condition="DRIVER_SBUS" /> + <fcode source="cgthree.fs" name="QEMU,cgthree.bin" condition="DRIVER_SBUS" /> <fcode source="vga.fs" name="QEMU,VGA.bin" condition="DRIVER_VGA" /> </build> Added: trunk/openbios-devel/drivers/cgthree.fs ============================================================================== --- /dev/null 00:00:00 1970 (empty, because file is newly added) +++ trunk/openbios-devel/drivers/cgthree.fs Mon Aug 19 14:58:09 2013 (r1222) @@ -0,0 +1,154 @@ +\ +\ Fcode payload for QEMU CG3 graphics card +\ +\ This is the Forth source for an Fcode payload to initialise +\ the QEMU CG3 graphics card. +\ +\ (C) Copyright 2013 Mark Cave-Ayland +\ + +fcode-version3 + +\ +\ Instead of using fixed values for the framebuffer address and the width +\ and height, grab the ones passed in by QEMU/generated by OpenBIOS +\ + +: (find-xt) \ ( str len -- xt | -1 ) + $find if + exit + else + -1 + then +; + +" openbios-video-width" (find-xt) cell+ value openbios-video-width-xt +" openbios-video-height" (find-xt) cell+ value openbios-video-height-xt +" depth-bits" (find-xt) cell+ value depth-bits-xt +" line-bytes" (find-xt) cell+ value line-bytes-xt +" debug-type" (find-xt) value debug-type-xt + +: openbios-video-width openbios-video-width-xt @ ; +: openbios-video-height openbios-video-height-xt @ ; +: depth-bits depth-bits-xt @ ; +: line-bytes line-bytes-xt @ ; +: debug-type debug-type-xt execute ; + +\ +\ Registers +\ + +h# 400000 constant cg3-off-dac +h# 20 constant /cg3-off-dac + +h# 800000 constant cg3-off-fb +h# c0000 constant /cg3-off-fb + +: >cg3-reg-spec ( offset size -- encoded-reg ) + >r 0 my-address d+ my-space encode-phys r> encode-int encode+ +; + +: cg3-reg + \ A real cg3 rom appears to just map the entire region with a + \ single entry + h# 0 h# 1000000 >cg3-reg-spec + " reg" property +; + +: do-map-in ( offset size -- virt ) + >r my-space r> " map-in" $call-parent +; + +: do-map-out ( virt size ) + " map-out" $call-parent +; + +\ +\ DAC +\ + +-1 value cg3-dac +-1 value fb-addr + +: dac! ( data reg# -- ) + cg3-dac + c! +; + +external + +: color! ( r g b c# -- ) + 0 dac! ( r g b ) + swap rot ( b g r ) + 4 dac! ( b g ) + 4 dac! ( b ) + 4 dac! ( ) +; + +headerless + +\ +\ Mapping +\ + +: dac-map + cg3-off-dac /cg3-off-dac do-map-in to cg3-dac +; + +: fb-map + cg3-off-fb h# c0000 do-map-in to fb-addr +; + +: map-regs + dac-map fb-map +; + +\ +\ Installation +\ + +" cgthree" device-name +" display" device-type +" SUNW,501-1415" model + +: qemu-cg3-driver-install ( -- ) + cg3-dac -1 = if + map-regs + fb-addr to frame-buffer-adr + default-font set-font + + frame-buffer-adr encode-int " address" property + + openbios-video-width openbios-video-height over char-width / over char-height / + fb8-install + then +; + +: qemu-cg3-driver-init + + cg3-reg + + openbios-video-height encode-int " height" property + openbios-video-width encode-int " width" property + line-bytes encode-int " linebytes" property + + h# 39 encode-int 0 encode-int encode+ " intr" property + + \ Monitor sense. Some searching suggests that this is + \ 5 for 1024x768 and 7 for 1152x900 + openbios-video-width h# 480 = if + h# 7 + else + h# 5 + then + encode-int " monitor-sense" property + + " SUNW" encode-string " manufacturer" property + " ISO8859-1" encode-string " character-set" property + h# c encode-int " cursorshift" property + + ['] qemu-cg3-driver-install is-install +; + +qemu-cg3-driver-init + +end0

1 0

[commit] r1221 - trunk/openbios-devel/arch/sparc32
by repository service Aug. 19, 2013

Aug. 19, 2013

Author: mcayland Date: Mon Aug 19 14:58:06 2013 New Revision: 1221 URL: http://tracker.coreboot.org/trac/openbios/changeset/1221 Log: SPARC32: handle data faults for peek/poke Signed-off-by: Bob Breuer <breuerr(a)mc.net> Signed-off-by: Mark Cave-Ayland <mark.cave-ayland(a)ilande.co.uk> Modified: trunk/openbios-devel/arch/sparc32/init.fs trunk/openbios-devel/arch/sparc32/lib.c Modified: trunk/openbios-devel/arch/sparc32/init.fs ============================================================================== --- trunk/openbios-devel/arch/sparc32/init.fs Mon Aug 19 14:58:03 2013 (r1220) +++ trunk/openbios-devel/arch/sparc32/init.fs Mon Aug 19 14:58:06 2013 (r1221) @@ -46,6 +46,36 @@ 0 ; +\ (peek) and (poke) implementation +defer sfsr@ +defer ignore-dfault + +:noname + \ ( addr xt -- false | value true ) + sfsr@ drop \ Clear any existing MMU fault status + + -1 ignore-dfault ! \ Disable data fault trap + execute + 0 ignore-dfault ! \ Enable data fault trap + + sfsr@ 0= if + true + else + drop false \ Failed, drop the read value + then +; to (peek) + +:noname + \ ( value addr xt -- okay? ) + sfsr@ drop \ Clear any existing MMU fault status + + -1 ignore-dfault ! \ Disable data fault trap + execute + 0 ignore-dfault ! \ Enable data fault trap + + sfsr@ 0= \ true if no fault +; to (poke) + \ Load TCX FCode driver blob [IFDEF] CONFIG_DRIVER_SBUS -1 value tcx-driver-fcode Modified: trunk/openbios-devel/arch/sparc32/lib.c ============================================================================== --- trunk/openbios-devel/arch/sparc32/lib.c Mon Aug 19 14:58:03 2013 (r1220) +++ trunk/openbios-devel/arch/sparc32/lib.c Mon Aug 19 14:58:06 2013 (r1221) @@ -302,6 +302,22 @@ DPRINTF("obp_dumb_memfree 0x%p (size %d)\n", va, sz); } +/* Data fault handling routines */ + +extern unsigned int ignore_dfault; + +/* ( -- reg ) */ +static void srmmu_get_sfsr(void) +{ + PUSH(srmmu_get_fstatus()); +} + +/* ( -- addr ) */ +static void ignore_dfault_addr(void) +{ + PUSH(pointer2cell(&ignore_dfault)); +} + void ob_init_mmu(void) { @@ -360,6 +376,12 @@ bind_func("pgmap@", pgmap_fetch); bind_func("pgmap!", pgmap_store); bind_func("map-pages", ob_map_pages); + + /* Install data fault handler words for cpeek etc. */ + PUSH_xt(bind_noname_func(srmmu_get_sfsr)); + feval("to sfsr@"); + PUSH_xt(bind_noname_func(ignore_dfault_addr)); + feval("to ignore-dfault"); } /*

1 0