October 2002 - coreboot - mail.coreboot.org

status
by Ronald G Minnich Oct. 18, 2002

Oct. 18, 2002

I have to do some other stuff, but: ar cr linuxbios.a crt0.o //home/rminnich/src/config/./normal/option_table.o linuxbiosmain.o linuxpci.o newpci.o clog2.o printk.o serial_subr.o subr.o vsprintf.o memset.o memcpy.o memcmp.o malloc.o elfboot.o do_inflate.o delay.o fallback_boot.o compute_ip_checksum.o version.o keyboard.o mc146818rtc.o isa-dma.o i8259.o ide.o boot.o linuxbios_table.o i386_subr.o params.o hardwaremain.o c_start.o pirq_routing.o mpspec.o ioapic.o start_stop.o secondary.o northbridge.o nvram.o ich3_ioapic.o ich3_lpc.o ich3_ide.o ich3_reset.o ich3_smbus.o ich3_cpu.o ich3_rtc.o ich3_power.o ich3_1e0_misc.o ich3_1f0_misc.o p64h2_ioapic.o p64h2_pcibridge.o superio_winbond_w83627hf.o nsuperio.o keyboard.o mc146818rtc.o isa-dma.o i8259.o ide.o generic_superio.o w83627hf_power.o power_led.o ramtest.o cpuid.o microcode.o mtrr.o l2_cache.o delay_i786.o mainboard.o mtrr_values.o mptable.o irq_tables.o gcc -nostdlib -r -o linuxbios.o crt0.o rom_fill_inbuf.o linuxbios.a /usr/lib/gcc-lib/i386-redhat-linux/3.2/libgcc.a linuxbios.a(c_start.o): In function `_start': c_start.o(.text+0x0): multiple definition of `_start' crt0.o(.rom.text+0x4): first defined here collect2: ld returned 1 exit status make: *** [linuxbios.o] Error 1 What's interesting: I have a build that works from eric in which _start is defined (as T) in crt0.0 and c_start.o; and I have this build, same thing. This build gets the error, the other build doesn't. More as a find it. ron

1 0

e7500 closer.
by Ronald G Minnich Oct. 17, 2002

Oct. 17, 2002

only remaining undefined's are gdt, gdt_limit, and udelay. If you're working with these boards, do an update. ron

1 0

E7500
by steven james Oct. 17, 2002

Oct. 17, 2002

Greetings, The last batch of commits helped. Now I'm getting undefined reference to SMBUS_MEM_CHANNEL_OFF doing find -exec grep, I don't see that defined anywhere. G'day, sjames -- -------------------------steven james, director of research, linux labs ... ........ ..... .... 230 peachtree st nw ste 701 the original linux labs atlanta.ga.us 30303 -since 1995 http://www.linuxlabs.com office 404.577.7747 fax 404.577.7743 -----------------------------------------------------------------------

2 1

gcc, ever improving ... :-)
by Ronald G Minnich Oct. 17, 2002

Oct. 17, 2002

printk_debug(__FUNCTION__ "\n"); You get the message: /home/rminnich/src/freebios/src/arch/i386/lib/hardwaremain.c:139: warning: concatenation of string literals with __FUNCTION__ is deprecated OK, I use this technique quite a bit. Anybody know why this handy little trick is 'deprecated'? ron

4 5

supermicro motherboards from the tree
by Ronald G Minnich Oct. 17, 2002

Oct. 17, 2002

I'm trying to build a supermicro e7500 motherboard from the sourceforge repo for testing. Currently I get: /home/rminnich/src/freebios/src/arch/i386/smp/secondary.inc Anyone else seen this? ron

2 2

e7500
by Ronald G Minnich Oct. 17, 2002

Oct. 17, 2002

Getting there. Eric has done a truly amazing amount of work ... Basic syntax errors and other compilation errors fixed. Now to shoot for link-time errors (e.g. udelay etc.) CVS: Modified Files: CVS: src/arch/i386/lib/Config src/include/pci.h CVS: src/include/pci_ids.h src/include/superio/w83627hf.h CVS: src/lib/newpci.c src/northbridge/intel/E7500/northbridge.c CVS: src/southbridge/intel/82801ca/ich3_ioapic.c CVS: src/southbridge/intel/82870/p64h2_ioapic.c ron

1 0

some commits ...
by Ronald G Minnich Oct. 17, 2002

Oct. 17, 2002

Simple fixes so that dependencies are satisfied; now for the compile errors. CVS: Modified Files: CVS: src/config/Config src/mainboard/supermicro/p4dpe/Config CVS: src/northbridge/intel/E7500/raminit.inc ~

1 0

mainboard status
by Ronald G Minnich Oct. 17, 2002

Oct. 17, 2002

Maintaining mainboard status has been a pain in the neck from the beginning, and as the number of mainboards has grown it has gotten harder and harder. Here is what I would like to do: for each src/mainboard/<vendor>/<board>, e.g. src/mainboard/tyan/guiness, I want to have a file called STATUS. In this file are keyword-value pairs describing the state of the mainboard. Each night at midnight we'll have a cron job collate this information and produce a web page. Mainboards will have owners responsible for letting us know if the board is working, and what the most recent good build time was. This information will be display on the web page. The file would look something like this: # These are keyword-value pairs. # a : separates the keyword from the value # the value is arbitrary text delimited by newline. # continuation, if needed, will be via the \ at the end of a line # comments are indicated by a '#' as the first character. # the keywords are case-INSENSITIVE owner: Ron Minnich email: rminnich(a)lanl.gov #status: One of unsupported, unstable, stable status: unstable # Last-known-good follows the internationl date standard: day/month/year last-known-good: 0/0/0000 Comments: Great mainboard for a dashboard radio Rather than have a single file with all the status, I want to have separate files for each mainboard so the various committers can work on their STATUS file without worrying about stomping on each other. Comments? Are there fields missing? Is there a fundamental problem with this idea that I'm missing? Thanks ron

6 14

Fwd: RE: mainboard status
by Antony Stone Oct. 17, 2002

Oct. 17, 2002

---------- Forwarded Message ---------- Subject: RE: mainboard status Date: Thu, 17 Oct 2002 16:11:32 -0400 From: "Todd E. Johnson" <tejohnson(a)yahoo.com> Wow, This sounds really cool. One quick question though... What about boards that are "Generic" in terms of labels, where you have board name A, and board name B, when they are in fact the same board? Would there be any benefit to a field called "OtherNames"? Or similar? Regards, Todd E. Johnson tejohnson(a)yahoo.com > -----Original Message----- > On Thursday 17 October 2002 5:59 pm, Ronald G Minnich wrote: > > I could just make a field called links? > > Sounds good. Then the individual mainboard owners can point > to whatever > information they like (whether it's theirs or someone else's). > > Antony. -- Most people have more than the average number of legs.

1 0

[ANNOUNCE] mkelfImage-1.17
by Eric W Biederman Oct. 17, 2002

Oct. 17, 2002

The differences in 2.5.42 weren't as bad as they looked at first glance. I just had to add some extra descriptors into the gdt. mkelfImage-1.17 is now released. You can find it at: ftp://ftp.lnxi.com/pub/src/mkelfImage/mkelfImage-1.17.tar.gz Eric

1 0