coreboot November 2004

coreboot@coreboot.org

31 participants
183 discussions

[RFC] linuxbios autobuild
by Stefan Reinauer 05 Nov '04

05 Nov '04

Hi, What's missing in the linuxbios autobuild tool? Regression Testing ------------------ We might want to regularly (or even on checkin?) use abuild.sh to check the LinuxBIOS repository for consistency and open building lots.. Shall we have a mail sent to the mailinglist/cvs list regularly/on checkin? We definitely also want a web page with this information, maybe the tool should optionally generate HTML/XML reports? ... ? Resulting ROM images -------------------- There needs to be a way to specify a per build or at least per platform payload. Otherwise no bootable images will result. Once the images have been proven stable we could offer prebuilt images (or have a webfrontend that creates an image from 2 payloads and 2 linuxbios romimages.. too lion-hearted? Development ----------- I would love to see the rest of the target config files go away for default builds, or have the abuild tool create them easily on the fly. This could be nice for testing different scenarios during development as well as for end users that don't want to go edit the default configuration but still want a different payload etc.. Ideas? What's missing? Stefan

2 1

Re: linuxbios autobuild / regression testing
by Stefan Reinauer 05 Nov '04

05 Nov '04

* Eric W. Biederman <ebiederman(a)lnxi.com> [041105 10:45]: > > It seems no mainboard says "use LD" at all yet. Some still don't > > "use CC" but this is minor.. > > Actually what you should be looking for is "uses CROSS_COMPILE" > Currently we don't seem to call LD directly so as long as gcc can > find the proper one we are ok. Oops. Misassumption on my side. I just saw the defines on the compile command line: -DLINUXBIOS_LINKER='"GNU ld version 2.14.90.0.5 20030722 (SuSE Linux)"' -DLINUXBIOS_ASSEMBLER='"GNU assembler version 2.14.90.0.5 (x86_64-suse-linux) using BFD version 2.14.90.0.5 20030722 (SuSE Linux)"' Are these compiled in per default? > > What's the usual naming convention for the cross compiler and cross > > linker of a given platform? We also need to cope with multiple payloads > > if we want to be able to use the resulting binaries... (in an ideal > > world the next step would be to automatically flash them) > > I suspect that we will want to have a simple Config.lb in targets that > we can start with and add cross compiling and other options to. Hm. We're setting a lot of default options in Options.lb and some in Config.lb for the mainboards.. Is there room to place defaults for these specifics in there? > > ./abuild.sh -at arima/hdama > -a and -t sound like useful additions. -at doesn't seem to work btw. -a -t does though.. whatever. > Just don't forget that ultimately if we can fix the dependencies in the > tree we both -a and my caching of what actually built successfully > should be trivial. so far I deleted the linuxbios-builds directory most of the time before each build to make sure nothing old is keeping things from building correctly because I saw a place or two where things did not get rebuild though something changed in the source tree. With the rebuild of all motherboards currently being around 10 minutes on this machine speed did not really care for generating the reports along. This is different if the tool is used actively in each cycle during development.. > Note: I had to tweak the parser so it would work when I specified where the > linuxbios tree was. It was putting single quotes around the pathname > nad ls was not really fond of './freebios2/'/src/mainboard .... Oh,.. I had that problem too with the -t option.. Thanks for fixing this. I will remove the stupid tr -d workaround I used. Stefan

1 0

Technologic ts5300???
by ebiederman＠lnxi.com 05 Nov '04

05 Nov '04

There is the beginnins of a port but there is no sc520 code. What is the status of this? With nothing checked into the tree I cannot fix it along with everything else. Eric

3 8

RE: Board hangs after soft_reset() in auto.c
by YhLu 05 Nov '04

05 Nov '04

Which stage auto.c E0: 0xff010003 --> 0x06000003 E4: 0x09070103 --> 0x09070103 E8: --> 0x0c0a0203 YH -----Original Message----- From: Liu Tao [mailto:liutao@safe-mail.net] Sent: Thursday, November 04, 2004 8:31 PM To: YhLu Cc: Eric W. Biederman; LinuxBIOS Subject: Re: Board hangs after soft_reset() in auto.c I mean when enumerating HT0 e0 is set to 0xff010003, while e4 matains 0x09070103, so e4 overlaps with e0? Regards, Liu Tao YhLu wrote: >That only used by auto.c (in_conherent.c) > >And it is ok to be overlapped in c parts. ( hardwaremain)? > >To do that, only let cpu can see different BUS in auto.c stage. >

2 1

make -j4 ?
by Stefan Reinauer 05 Nov '04

05 Nov '04

Hi, just noticed on my way to bed.. it seems make -j does work fine for the single linuxbios images (fallback / normal) but not for the concatenated final one because of a wrong dependency: ------------------------------ 8< ----------------------------- all: fallback-rom normal-rom buildroms fallback-rom: if (cd fallback; \ make linuxbios.rom)\ then true; else exit 1; fi; normal-rom: if (cd normal; \ make linuxbios.rom)\ then true; else exit 1; fi; buildroms: cat normal/linuxbios.rom fallback/linuxbios.rom > ./amd_quartet.rom ------------------------------ 8< ----------------------------- should look like: all: buildroms [..] buildroms: fallback-rom normal-rom Otherwise make assumes it can build all three of fallback-rom, normal-rom and buildroms in parallel Stefan

2 1

RE: Removing device driver code in mainboard.c for Tyan boards?
by YhLu 04 Nov '04

04 Nov '04

No, debug_dev can be PCI_DOMAIN and APIC's sibling and one child dev_root in link0. YH -----Original Message----- From: ebiederman(a)lnxi.com [mailto:ebiederman@lnxi.com] Sent: Thursday, November 04, 2004 8:29 PM To: YhLu Cc: Ronald G. Minnich; Li-Ta Lo; LinuxBIOS Subject: Re: Removing device driver code in mainboard.c for Tyan boards? YhLu <YhLu(a)tyan.com> writes: > Great. That will be very clear. > > Then > chip northbridge/amd/amk8/root_complex > > .. > > End # ----> dev_root link0 > chip northbridge/amd/amk8/root_complex > > .. > > End # ----> dev_root link1 > > chip northbridge/amd/amk8/root_complex > > .. > > End # ----> dev_root link2 And what do you want on link1 and link2 for the root_device? Among other things I'm not certain config.g currently supports that. Is this just for placing the debug devices far away? Eric

2 1

RE: Removing device driver code in mainboard.c for Tyan boards?
by YhLu 04 Nov '04

04 Nov '04

Great. That will be very clear. Then chip northbridge/amd/amk8/root_complex .. End # ----> dev_root link0 chip northbridge/amd/amk8/root_complex .. End # ----> dev_root link1 chip northbridge/amd/amk8/root_complex .. End # ----> dev_root link2

2 1

RE: Board hangs after soft_reset() in auto.c
by YhLu 04 Nov '04

04 Nov '04

That only used by auto.c (in_conherent.c) And it is ok to be overlapped in c parts. ( hardwaremain)? To do that, only let cpu can see different BUS in auto.c stage. Regards YH -----Original Message----- From: Liu Tao [mailto:liutao@safe-mail.net] Sent: Thursday, November 04, 2004 8:05 PM To: YhLu Cc: Eric W. Biederman; LinuxBIOS Subject: Re: Board hangs after soft_reset() in auto.c Can this cause configuration regions overlap with each other during device enumeration? In amdk8_scan_chains() the subordinate/limit field of current register is set to 0xff before hypertransport_scan_chain(), while other following registers are not touched yet, so they overlay with current register? Regards, Liu Tao > > PCI_ADDR(0, 0x18, 1, 0xE0), 0x0000FC88, 0x06000203, // AMD 8131/8111 on >link0 of CPU 0 > PCI_ADDR(0, 0x18, 1, 0xE4), 0x0000FC88, 0x09070003, // AMD 8131 on link1 >of CPU 0 > PCI_ADDR(0, 0x18, 1, 0xE8), 0x0000FC88, 0x0c0a0003, // AMD 8131 on link2 >of CPU 0 > PCI_ADDR(0, 0x18, 1, 0xEC), 0x0000FC88, 0x00000000, >

2 1

RE: Removing device driver code in mainboard.c for Tyan boards?
by YhLu 04 Nov '04

04 Nov '04

and Mainboard on ..... End Mainboard on ... End To put them into dev_root links YH -----Original Message----- From: YhLu Sent: Thursday, November 04, 2004 8:01 PM To: ebiederman(a)lnxi.com Cc: Ronald G. Minnich; Li-Ta Lo; LinuxBIOS Subject: RE: Removing device driver code in mainboard.c for Tyan boards? chip northbridge/amd/amdk8 device pci_domain 0 on device pci 18.0 on # LDT0 . .... End .... chip northbridge/amd/amdk8 device pci 19.0 end ... End End # PCI domain .....acpic... End # NB K8 It looks strang, Why not thang that to followings? device pci_domain 0 on chip northbridge/amd/amdk8 device pci 18.0 on # LDT0 . .... End .... End #NB K8 chip northbridge/amd/amdk8 device pci 19.0 end ... End End # PCI domain .....apic.... _______________________________________________ Linuxbios mailing list Linuxbios(a)clustermatic.org http://www.clustermatic.org/mailman/listinfo/linuxbios

2 1

RE: Removing device driver code in mainboard.c for Tyan boards?
by YhLu 04 Nov '04

04 Nov '04

chip northbridge/amd/amdk8 device pci_domain 0 on device pci 18.0 on # LDT0 . .... End .... chip northbridge/amd/amdk8 device pci 19.0 end ... End End # PCI domain .....acpic... End # NB K8 It looks strang, Why not thang that to followings? device pci_domain 0 on chip northbridge/amd/amdk8 device pci 18.0 on # LDT0 . .... End .... End #NB K8 chip northbridge/amd/amdk8 device pci 19.0 end ... End End # PCI domain .....apic....

2 1

← Newer
1
...
9
10
11
12
13
14
15
...
19
Older →

Jump to page:

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

coreboot November 2004