January 2016 - OpenBIOS - mail.coreboot.org

[PATCH 0/5] Enable bus master support for rtl8139 network cards on Apple PPC machines
by Mark Cave-Ayland Aug. 7, 2016

Aug. 7, 2016

This patchset is loosely based on earlier work by BALATON Zoltan <balaton(a)eik.bme.hu> and provides basic infrastructure to allow OpenBIOS to enable PCI device bus mastering. Following on from discussions on the QEMU mailing list, it seems that Apple's OF enables bus mastering for some PCI devices by default, and as a result some buggy drivers forget to explicitly enable it and hence these devices fail under QEMU's emulation. The first 3 patches add the basic support routines while the last 2 patches enable bus mastering for the rtl8139 card on Apple PPC machines which is required for OS X and MorphOS. Signed-off-by: Mark Cave-Ayland <mark.cave-ayland(a)ilande.co.uk> Mark Cave-Ayland (5): pci: introduce ob_pci_enable_bus_master() function pci: introduce ob_pci_is_bus_master_capable() function pci: add PCI database entry for rtl8139 network card pci: add rtl8139_config_cb() to configure rtl8139 network cards ppc: mark PCI slots 0-2 for Apple PPC machies as bus master capable openbios-devel/arch/ppc/qemu/init.c | 9 ++++--- openbios-devel/drivers/pci.c | 44 +++++++++++++++++++++++++++++++++ openbios-devel/drivers/pci_database.c | 6 +++++ openbios-devel/drivers/pci_database.h | 4 +++ openbios-devel/include/drivers/pci.h | 2 ++ 5 files changed, 62 insertions(+), 3 deletions(-) -- 1.7.10.4

5 17

[PATCH] ppc: force target CPU
by Laurent Vivier Feb. 2, 2016

Feb. 2, 2016

With gcc-5.2 default target is power7. As openbios targets PowerMacs G3/G4/G5, force gcc target to be G3 (32bit) or G5 (64bit). Signed-off-by: Laurent Vivier <laurent(a)vivier.eu> --- config/scripts/switch-arch | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/config/scripts/switch-arch b/config/scripts/switch-arch index 38d263d..bcf807a 100755 --- a/config/scripts/switch-arch +++ b/config/scripts/switch-arch @@ -265,7 +265,7 @@ for ARCH in $arch_list; do ppc) select_prefix powerpc powerpc64 if [ "$unix" = "no" ]; then - CFLAGS="-m32 -msoft-float -fno-builtin-bcopy -fno-builtin-log2" + CFLAGS="-m32 -mcpu=G3 -msoft-float -fno-builtin-bcopy -fno-builtin-log2" AS_FLAGS="-m32" else CFLAGS="-fno-builtin" @@ -275,7 +275,7 @@ for ARCH in $arch_list; do ppc64) select_prefix powerpc64 - CFLAGS="-Wa,-a64 -m64 -msoft-float -fno-builtin" + CFLAGS="-Wa,-a64 -m64 -mcpu=G5 -msoft-float -fno-builtin" AS_FLAGS="-Wa,-a64" ;; -- 2.5.0

8 35

Re: [OpenBIOS] [PATCH 1/1] [FCODE-utils] Allow overriding more Makefile vars
by Mark Cave-Ayland Jan. 21, 2016

Jan. 21, 2016

On 21/01/16 14:12, Peter Lemenkov wrote: > Hello All! > Sorry for the delayed answer. > > 2016-01-18 14:53 GMT+01:00 Segher Boessenkool <segher(a)kernel.crashing.org>: > >> Why not? Manual says to do just this; and I tested it just now to make >> sure I am not imagining things, and it works fine. So what is unusual >> in your setup? > > I accidentally confused you - everything works fine if I pass > variables like you suggested. But I'd like to add possibility to > override variables using envronment variables. > > This is exactly how various Linux distros' buildsystems are working. > So being able to setup variables system-wide would be helpful! If you're trying to override DESTDIR then you'd already need something that executes "make DESTDIR=/path/to/destdir" so if this works as Segher suggests then can you not do "make DESTDIR=$DESTDIR" or similar? ATB, Mark.

1 0

Re: [OpenBIOS] [PATCH 1/1] [FCODE-utils] Allow overriding more Makefile vars
by Segher Boessenkool Jan. 21, 2016

Jan. 21, 2016

On Mon, Jan 18, 2016 at 01:38:04PM +0100, Peter Lemenkov wrote: > >> > But why do you need it? Why doesn't e.g. > >> > > >> > make DESTDIR=/opt/bla > >> > > >> > work for you as-is? Or do you need those vars to be passed in via the > >> > environment? > >> > >> Unfortunately this DESTDIR trick won't work. Explicitly stated > >> variables from the Makefile itself have preference over the > >> environment variables unless they stated with "?=". At least this is > >> how GNU make works. > > > > But this is *not* an environment variable! > > Yes, sorry. Anyway we can't override Makefile-variables that way > unless we allow it. Why not? Manual says to do just this; and I tested it just now to make sure I am not imagining things, and it works fine. So what is unusual in your setup? Segher

2 1

Re: [OpenBIOS] [PATCH 1/1] [FCODE-utils] Allow overriding more Makefile vars
by Segher Boessenkool Jan. 18, 2016

Jan. 18, 2016

On Mon, Jan 18, 2016 at 11:04:34AM +0100, Peter Lemenkov wrote: > Hello All! > > 2016-01-18 7:50 GMT+01:00 Segher Boessenkool <segher(a)kernel.crashing.org>: > > > It probably works fine. "?=" can be problematic with e.g. recursive > > makefiles, but you don't have those here. > > Also it's already used within the Makefiles to override C-compiler, so > my patch won't change the current situation. > > > But why do you need it? Why doesn't e.g. > > > > make DESTDIR=/opt/bla > > > > work for you as-is? Or do you need those vars to be passed in via the > > environment? > > Unfortunately this DESTDIR trick won't work. Explicitly stated > variables from the Makefile itself have preference over the > environment variables unless they stated with "?=". At least this is > how GNU make works. But this is *not* an environment variable! Segher

2 1

[PATCH 1/1] [FCODE-utils] Allow overriding more Makefile vars
by Peter Lemenkov Jan. 18, 2016

Jan. 18, 2016

As a package maintainer I'd like to override some more Makefile variables. Namely: * DESTDIR * strip command invokation (ability to suppress it by passing /bin/true instead of /bin/strip) * pass custom gcc flags This patch allows this. It does not break any existing workflows where it's not required to change these variables. Signed-off-by: Peter Lemenkov <lemenkov(a)gmail.com> --- detok/Makefile | 6 +++--- romheaders/Makefile | 6 +++--- toke/Makefile | 6 +++--- 3 files changed, 9 insertions(+), 9 deletions(-) diff --git a/detok/Makefile b/detok/Makefile index 3ba6c30..0e422fb 100644 --- a/detok/Makefile +++ b/detok/Makefile @@ -24,13 +24,13 @@ PROGRAM = detok -DESTDIR = /usr/local +DESTDIR ?= /usr/local CC ?= gcc -STRIP = strip +STRIP ?= strip INCLUDES = -I../shared # Normal Flags: -CFLAGS = -O2 -Wall #-Wextra +CFLAGS ?= -O2 -Wall #-Wextra LDFLAGS = # Coverage: diff --git a/romheaders/Makefile b/romheaders/Makefile index a216e7b..1831036 100644 --- a/romheaders/Makefile +++ b/romheaders/Makefile @@ -24,10 +24,10 @@ PROGRAM = romheaders -DESTDIR = /usr/local +DESTDIR ?= /usr/local CC ?= gcc -STRIP = strip -CFLAGS = -O2 -Wall -Wextra +STRIP ?= strip +CFLAGS ?= -O2 -Wall -Wextra INCLUDES = -I../shared SOURCES = romheaders.c ../shared/classcodes.c diff --git a/toke/Makefile b/toke/Makefile index aa75acd..8e31abd 100644 --- a/toke/Makefile +++ b/toke/Makefile @@ -24,13 +24,13 @@ PROGRAM = toke -DESTDIR = /usr/local +DESTDIR ?= /usr/local CC ?= gcc -STRIP = strip +STRIP ?= strip INCLUDES = -I../shared # Normal flags -CFLAGS = -O2 -Wall #-Wextra +CFLAGS ?= -O2 -Wall #-Wextra LDFLAGS = # Coverage: -- 2.5.0

3 3

Progress on Sparc64?
by Nick Couchman Jan. 17, 2016

Jan. 17, 2016

Wondering if there have been any recent developments getting OpenBIOS + Qemu working for the Sparc64 platform? I have a couple of legacy applications that require 64-bit Sparc support that would nice to have running on here. Anything I can do to help debug issues? I have ISOs for 64-bit Solaris 8, 9, 10, and 11, and am happy to test stuff out. -Nick == This e-mail may contain SEAKR Engineering (SEAKR) Confidential and Proprietary Information. If this message is not intended for you, you are strictly prohibited from using this message, its contents or attachments in any way. If you have received this message in error, please delete the message from your mailbox. This e-mail may contain export-controlled material and should be handled accordingly.

2 2

Re: [OpenBIOS] Progress on Sparc64?
by Nick Couchman Jan. 16, 2016

Jan. 16, 2016

----- Original Message ----- > From: "Programmingkid" <teammember0x01(a)gmail.com> > To: "Nick E Couchman" <nick.couchman(a)seakr.com> > Cc: openbios(a)openbios.org > Sent: Friday, January 15, 2016 10:46:57 PM > Subject: Re: [OpenBIOS] Progress on Sparc64? > On Jan 16, 2016, at 12:32 AM, Nick Couchman wrote: > >> Wondering if there have been any recent developments getting OpenBIOS + Qemu >> working for the Sparc64 platform? I have a couple of legacy applications that >> require 64-bit Sparc support that would nice to have running on here. Anything >> I can do to help debug issues? I have ISOs for 64-bit Solaris 8, 9, 10, and >> 11, and am happy to test stuff out. >> >> -Nick > > http://wiki.qemu.org/SPARC > This page has a table that would benefit from your testing data. I only saw > Solaris 1 and 8 on it. We would welcome Solaris 9, 10, and 11 to the table. Okay. Any chance you have a Wiki account and can you create an account for me so that I can edit the page? Below are some of the results that I'm seeing trying to boot various versions of Solaris on the qemu-system-sparc64 platform. Let me know if there is other debugging info that would be useful - different qemu switches, gdb output, and/or Solaris boot parameters. ==Solaris 8 64-bit on Sparc64== [root@qemu ~]# !68 /opt/qemu/bin/qemu-system-sparc64 -nographic -cdrom /root/sol-8-hw4-install-sparc.iso -boot d -m 2048M OpenBIOS for Sparc64 Configuration device id QEMU version 1 machine id 0 kernel cmdline CPUs: 1 x SUNW,UltraSPARC-IIi UUID: 00000000-0000-0000-0000-000000000000 Welcome to OpenBIOS v1.1 built on Oct 27 2015 23:43 Type 'help' for detailed information Trying cdrom:f... Not a bootable ELF image Not a bootable a.out image Loading FCode image... Loaded 5936 bytes entry point is 0x4000 open isn't unique. Jumping to entry point 0000000000100000 for type 0000000000000001... switching to new context: entry point 0x100000 stack 0x00000000ffe8aa09 warning:interpret: exception -13 caught interpret warning @ warning off : sync %tl-c %tstate h# 104185c0 x! %g1 h# 104185c8 x! %g2 h# 104185d0 x! %g3 h# 104185d8 x! %g4 h# 104185e0 x! %g5 h# 104185e8 x! %g6 h# 104185f0 x! %g7 h# 104185f8 x! %o0 h# 10418600 x! %o1 h# 10418608 x! %o2 h# 10418610 x! %o3 h# 10418618 x! %o4 h# 10418620 x! %o5 h# 10418628 x! %o6 h# 10418630 x! %o7 h# 10418638 x! %tl-c %tpc h# 10418640 x! %tl-c %tnpc h# 10418648 x! %y h# 10418650 l! %tl-c %tt h# 104185b8 x! sync ; warning ! failed with error ffffffffffffffed SunOS Release 5.8 Version Generic_108528-29 64-bit Copyright 1983-2003 Sun Microsystems, Inc. All rights reserved. spacex@:interpret: exception -13 caught interpret h# d constant MMU_PAGESHIFT h# 0 constant TTE8K h# 20 constant SFHME_SIZE h# 0 constant SFHME_TTE h# 8 constant HMEBLK_TAG h# 0 constant HMEBLK_NEXT h# 2c constant HMEBLK_MISC h# 38 constant HMEBLK_HME1 h# 8 constant NHMENTS h# 3 constant HBLK_SZMASK h# 10 constant HBLK_RANGE_SHIFT h# 8 constant HMEBP_HBLK h# 20 constant HMEBUCKET_SIZE h# 0 constant HTAG_SFMMUPSZ h# d constant HTAG_REHASHSZ h# 3 constant MAX_HASHCNT h# 7c900000 constant uhme_hash h# 7c800000 constant khme_hash h# 1ffff constant U\ panic[cpu0]/thread=10408000: boot_alloc: pp is NULL or free 00000000104071d0 unix:boot_mapin+7c (2000, 1041b2d0, 300001fa000, 0, 3000020e000, 0) %l0-3: 000003000020c000 ffffffffffffe000 000000001044aef8 0000000000014000 %l4-7: 0000030fffff7f10 0000000000000000 000000001040d970 0000000000007fa3 0000000010407280 unix:boot_alloc+50 (14000, 14000, 1000, 300001fa000, 1000c368, 0) %l0-3: 0000000010026090 0000000010407428 000000001040e000 0000000000000000 %l4-7: 0000000000000016 0000000000000000 0000000000000016 0000000000000000 0000000010407330 unix:segkmem_alloc+30 (300001fa000, 14000, 0, 0, 10451c98, 1044b300) %l0-3: 0000030fffff4268 00000300001ae390 0000000000000000 000000001044bdc0 %l4-7: 000000001044af18 000000001044aef8 0000000010452498 0000000000000000 00000000104073f0 genunix:vmem_xalloc+3e4 (10451670, 10451a78, ffffffffffffffff, ffffffffffffffff, 0, 0) %l0-3: 000000001004a3b4 ffffffffffffe000 0000000010451670 0000000000014000 %l4-7: 0000000000000000 0000000000000000 0000000000014000 0000000010451690 0000000010407520 genunix:vmem_alloc+34 (10451670, 14000, 0, 10438978, 1041c6a8, 4e0) %l0-3: 0000000010438970 00000300000f7778 0000000000000000 0000000000000001 %l4-7: 0000000000000001 0000000010438978 000000000000001a 0000000000000001 00000000104075e0 genunix:kmem_alloc+b4 (14000, 0, 14000, 1, f0, 0) %l0-3: 00000300000f7a00 ffffffffffffffc0 000000001044f8e0 00000000000003c0 %l4-7: 0000030fffff7378 0000000000000000 0000000000000016 00000000104075f3 0000000010407690 genunix:kmem_zalloc+a0 (14000, 0, 104542c8, 0, 0, 300000f7680) %l0-3: 000000001043a0b0 0000000000000028 00000300000f7a00 0000000000000008 %l4-7: ffffffffffffffff 0000000000000001 0000000000000000 0000000000000000 0000000010407740 genunix:impl_ddi_callback_init+98 (0, 1003f354, 0, 0, 0, 10407897) %l0-3: 0000000010444c00 0000000010419400 0000000000000000 0000000000000001 %l4-7: 0000000000000001 0000000010439720 0000000000000002 0000000000000001 0000000010407810 unix:setup_ddi+24 (0, 10424c50, 1041b2c8, 0, 0, 10407897) %l0-3: 0000000010114380 000000001013c798 000000007c900000 0000000000012000 %l4-7: 00000000000000b0 0000000010413c48 000000001040d970 0000000000007fa3 00000000104078c0 unix:startup_modules+1e8 (1040e000, 10414ab8, 30000000000, 10418668, 2000, 30000016000) %l0-3: 0000000000000103 0000000010423c00 0000000000000020 0000031002c00000 %l4-7: 0000000010418400 000000001041d000 000000001041c000 000000001041b000 0000000010407970 unix:startup+14 (10428c00, 0, 0, 1, 0, ffffffffffffffff) %l0-3: 0000000010026090 000000000000d925 0000000000000afd 0000000000000000 %l4-7: 0000000010472880 00000000002f0c43 00000000000beafd 0000000000000afd 0000000010407a20 genunix:main+4 (1040d400, 2000, 10407ec0, 10408030, fff2, 10052a0c) %l0-3: 0000000010408000 0000000000000001 0000000000000015 0000000000000f36 %l4-7: 0000000010429618 0000000010472880 00000000000d7438 0000000000000540 skipping system dump - no dump device configured rebooting... BOOTpanic - kernel: prom_reboot: reboot call returned! EXIT 0 > == End Solaris 8 64-bit == == Solaris 9 on Sparc64 == [root@qemu ~]# /opt/qemu/bin/qemu-system-sparc64 -nographic -cdrom /root/sol-9-905hw-ga-sparc-dvd.iso -boot d -m 2048M OpenBIOS for Sparc64 Configuration device id QEMU version 1 machine id 0 kernel cmdline CPUs: 1 x SUNW,UltraSPARC-IIi UUID: 00000000-0000-0000-0000-000000000000 Welcome to OpenBIOS v1.1 built on Oct 27 2015 23:43 Type 'help' for detailed information Trying cdrom:f... Not a bootable ELF image Not a bootable a.out image Loading FCode image... Loaded 5936 bytes entry point is 0x4000 open isn't unique. Jumping to entry point 0000000000100000 for type 0000000000000001... switching to new context: entry point 0x100000 stack 0x00000000ffe8aa09 warning:interpret: exception -13 caught interpret warning @ warning off : sync %tl-c %tstate h# 1418608 x! %g1 h# 1418610 x! %g2 h# 1418618 x! %g3 h# 1418620 x! %g4 h# 1418628 x! %g5 h# 1418630 x! %g6 h# 1418638 x! %g7 h# 1418640 x! %o0 h# 1418648 x! %o1 h# 1418650 x! %o2 h# 1418658 x! %o3 h# 1418660 x! %o4 h# 1418668 x! %o5 h# 1418670 x! %o6 h# 1418678 x! %o7 h# 1418680 x! %tl-c %tpc h# 1418688 x! %tl-c %tnpc h# 1418690 x! %y h# 1418698 l! %tl-c %tt h# 1418600 x! sync ; warning ! failed with error ffffffffffffffed SunOS Release 5.9 Version Generic_118558-34 64-bit Copyright 1983-2003 Sun Microsystems, Inc. All rights reserved. Use is subject to license terms. - panic[cpu0]/thread=140a000: boot_alloc: pp is NULL or free 0000000001409120 unix:boot_mapin+80 (30000146000, 82000, 82000, 1000, 0, 30000004020) %l0-3: 00000300001c6000 00000300001c8000 0000000000000000 000000000142b4f8 %l4-7: 0000000000002000 0000000000000000 ffffffffffffffff ffffffffffffffff 00000000014091d0 unix:boot_alloc+4c (30000146000, 82000, 1000, 1422aa8, 0, 0) %l0-3: 0000000000000002 0000000000082000 000000000146bc00 0000000000081fff %l4-7: 0000030000004020 0000000000000000 ffffffffffffffff ffffffffffffffff 0000000001409280 unix:segkmem_alloc+8c (30000004000, 82000, 0, 0, 0, 0) %l0-3: 0000030000146000 0000000000082000 000000000146bc00 0000000001479360 %l4-7: 000003000003d7d8 0000000000000000 000000000116d398 00000000014229a8 0000000001409340 genunix:vmem_xalloc+4cc (ffffffffffffe000, 300000062d0, 0, 0, 0, 30000006020) %l0-3: 0000000000000004 000000000104fa18 0000030000006000 0000000000000000 %l4-7: 0000000000080008 0000000000000000 ffffffffffffffff ffffffffffffffff 0000000001409470 genunix:vmem_alloc+1f4 (30000006000, 80008, 0, 1422a1e, 144dcfb, 1422a00) %l0-3: 0000000000000000 0000000001422778 0000000000000008 0000000000080007 %l4-7: 0000030000006020 0000000000002000 000000000116d398 0000000000000000 0000000001409530 genunix:kmem_alloc+f0 (80008, 0, 20, 1a, 4e0, 1) %l0-3: 0000030000087640 000000000144f800 0000000000000000 0000030000087640 %l4-7: 000000000140d800 0000000001429ae0 0000000001429ab8 000000000144dcf0 00000000014095e0 genunix:kmem_zalloc+e4 (80008, 0, 4, a, 0, 0) %l0-3: 000000000144f8e0 0000030000087448 000000000140d800 ffffffffffffffff %l4-7: 0000000000000001 0000030000087640 0000000000000000 000000000144dcb8 0000000001409690 genunix:init_actions+38 (0, ffffffffffffffff, 5c, 0, 100c7e4, 0) %l0-3: 000000000144f400 0000000001444000 000000000144f6c8 0000000000000000 %l4-7: 0000000000000000 ffffffffffffffff 0000000000000001 0000030000066198 0000000001409760 genunix:ipp_init+c (100, 149d9e8, 1000, 12020, 78004000, 30000084db0) %l0-3: 000000000116cad4 000000000144d810 000000000149d048 00000000014439e8 %l4-7: 0000000000000001 0000000000000018 00000000014439f8 000000000149d000 0000000001409810 genunix:mod_setup+f4 (0, 1413240, 0, 78004000, 100, 78004000) %l0-3: 0000000000000000 0000000001497c00 0000000001495800 000000000000fc00 %l4-7: 0000000001077800 0000000078004000 000000007801601f ffffffffffffffff 00000000014098c0 unix:startup_modules+78 (12000, 142b000, 30000018000, 103, 78004000, 1432000) %l0-3: 0000000000000103 0000000000000000 000000000140c000 00000000014186b0 %l4-7: 00000000014186c0 0000000078004000 0000000000002000 0000000000012020 0000000001409970 unix:startup+1c (1, 0, 0, 142e6b8, 1400000, 0) %l0-3: 0000000000000001 0000000000000000 00000000babecafe 0000000000000000 %l4-7: 000000000142c000 000000000140e000 0000000001416208 0000000001439ca8 0000000001409a20 genunix:main+4 (1409ba0, ffd0cbb0, 1409ec0, 383cbf, 2000, 500) %l0-3: 000000000140b400 000000000140a000 0000000001413020 0000000000000000 %l4-7: 0000000078002000 000000000038c000 00000000014a0198 0000000001067d98 skipping system dump - no dump device configured rebooting... BOOTpanic - kernel: prom_reboot: reboot call returned! EXIT 0 > == End Solaris 9 on Sparc64 == == Solaris 10 on Sparc64 == [root@qemu ~]# /opt/qemu/bin/qemu-system-sparc64 -nographic -cdrom /root/sol-10-u11-ga-sparc-dvd.iso -boot d -m 2048M -cpu "Sun UltraSparc IIIi" OpenBIOS for Sparc64 Configuration device id QEMU version 1 machine id 0 kernel cmdline CPUs: 1 x SUNW,UltraSPARC-IIIi UUID: 00000000-0000-0000-0000-000000000000 Welcome to OpenBIOS v1.1 built on Oct 27 2015 23:43 Type 'help' for detailed information Trying cdrom:f... Not a bootable ELF image Not a bootable a.out image Loading FCode image... Loaded 7420 bytes entry point is 0x4000 == End Solaris 10 on Sparc64 == == This e-mail may contain SEAKR Engineering (SEAKR) Confidential and Proprietary Information. If this message is not intended for you, you are strictly prohibited from using this message, its contents or attachments in any way. If you have received this message in error, please delete the message from your mailbox. This e-mail may contain export-controlled material and should be handled accordingly.

2 1

[commit] r1377 - trunk/openbios-devel/drivers
by repository service Jan. 8, 2016

Jan. 8, 2016

Author: mcayland Date: Fri Jan 8 13:09:48 2016 New Revision: 1377 URL: http://tracker.coreboot.org/trac/openbios/changeset/1377 Log: pci: don't hard-code PCI interrupt maps by default The current code for openpic interrupts is incorrect for two reasons: firstly it generates an interrupt-map property with 8 IRQ pins (when only 4 are specified and so the last 4 entries are junk), and secondly it hard-codes the entries with a non-standard interrupt-mask to give one entry per pin rather than one entry per PCI device. Switch everything over to use the existing code used by SPARC64 which dynamically enumerates the child nodes and builds up the correct properties on demand. This ensures that everything will work on PCI platforms particularly if extra cards are dynamically plugged into QEMU. Signed-off-by: Mark Cave-Ayland <mark.cave-ayland(a)ilande.co.uk> Modified: trunk/openbios-devel/drivers/pci.c Modified: trunk/openbios-devel/drivers/pci.c ============================================================================== --- trunk/openbios-devel/drivers/pci.c Fri Jan 8 13:09:46 2016 (r1376) +++ trunk/openbios-devel/drivers/pci.c Fri Jan 8 13:09:48 2016 (r1377) @@ -1428,9 +1428,11 @@ static void ob_pci_host_set_interrupt_map(phandle_t host) { - phandle_t dnode = 0; - u32 props[128]; - int i; + phandle_t dnode = 0, pci_childnode = 0; + u32 props[128], intno; + int i, ncells, len; + u32 *val, addr; + char *reg; #if defined(CONFIG_PPC) phandle_t target_node; @@ -1474,69 +1476,60 @@ target_node = find_dev("/pci"); set_int_property(target_node, "interrupt-parent", dnode); - - /* openpic interrupt mapping */ - for (i = 0; i < (7*8); i += 7) { - props[i + PCI_INT_MAP_PCI0] = 0; - props[i + PCI_INT_MAP_PCI1] = 0; - props[i + PCI_INT_MAP_PCI2] = 0; - props[i + PCI_INT_MAP_PCI_INT] = (i / 7) + 1; // starts at PINA=1 - props[i + PCI_INT_MAP_PIC_HANDLE] = dnode; - props[i + PCI_INT_MAP_PIC_INT] = arch->irqs[i / 7]; - props[i + PCI_INT_MAP_PIC_POL] = 3; - } - set_property(host, "interrupt-map", (char *)props, 7 * 8 * sizeof(props[0])); - - props[PCI_INT_MAP_PCI0] = 0; - props[PCI_INT_MAP_PCI1] = 0; - props[PCI_INT_MAP_PCI2] = 0; - props[PCI_INT_MAP_PCI_INT] = 0x7; - - set_property(host, "interrupt-map-mask", (char *)props, 4 * sizeof(props[0])); } -#elif defined(CONFIG_SPARC64) - int ncells, len; - u32 *val, addr; - char *reg; +#else + /* PCI host bridge is the default interrupt controller */ + dnode = host; +#endif /* Set interrupt-map for PCI devices with an interrupt pin present */ ncells = 0; PUSH(host); fword("child"); - dnode = POP(); - while (dnode) { - if (get_int_property(dnode, "interrupts", &len)) { - reg = get_property(dnode, "reg", &len); - if (reg) { + pci_childnode = POP(); + while (pci_childnode) { + intno = get_int_property(pci_childnode, "interrupts", &len); + if (len && intno) { + reg = get_property(pci_childnode, "reg", &len); + if (len && reg) { val = (u32 *)reg; for (i = 0; i < (len / sizeof(u32)); i += 5) { addr = val[i]; /* Device address is in 1st 32-bit word of encoded PCI address for config space */ - if (!(addr & 0x03000000)) { + if ((addr & PCI_RANGE_TYPE_MASK) == PCI_RANGE_CONFIG) { +#ifdef CONFIG_SPARC64 ncells += pci_encode_phys_addr(props + ncells, 0, 0, addr, 0, 0); - props[ncells++] = 1; /* always interrupt pin 1 for QEMU */ - props[ncells++] = host; - props[ncells++] = SUN4U_INTERRUPT(addr, 1); + props[ncells++] = intno; + props[ncells++] = dnode; + props[ncells++] = SUN4U_INTERRUPT(addr, intno); +#endif + +#ifdef CONFIG_PPC + ncells += pci_encode_phys_addr(props + ncells, 0, 0, addr, 0, 0); + props[ncells++] = intno; + props[ncells++] = dnode; + props[ncells++] = arch->irqs[intno - 1]; + props[ncells++] = 3; +#endif } } } } - PUSH(dnode); + PUSH(pci_childnode); fword("peer"); - dnode = POP(); + pci_childnode = POP(); } set_property(host, "interrupt-map", (char *)props, ncells * sizeof(props[0])); props[0] = 0x0000f800; props[1] = 0x0; props[2] = 0x0; - props[3] = 7; + props[3] = 0x7; set_property(host, "interrupt-map-mask", (char *)props, 4 * sizeof(props[0])); -#endif } int ob_pci_init(void)

1 0

[commit] r1376 - trunk/openbios-devel/drivers
by repository service Jan. 8, 2016

Jan. 8, 2016

Author: mcayland Date: Fri Jan 8 13:09:46 2016 New Revision: 1376 URL: http://tracker.coreboot.org/trac/openbios/changeset/1376 Log: pci: remove the configuration space range from the PCI host bridge by default While the configuration space range appears in real SPARC device trees, it isn't mentioned in the IEEE-1275 PCI bindings and in fact causes Darwin/OS X to calculate PCI address spaces incorrectly. Disable this range in the PCI host bridge by default, except for SPARC64 where it can evidently still appear. Signed-off-by: Mark Cave-Ayland <mark.cave-ayland(a)ilande.co.uk> Reviewed-by: Alexander Graf <agraf(a)suse.de> Signed-off-by: Mark Cave-Ayland <mark.cave-ayland(a)ilande.co.uk> Modified: trunk/openbios-devel/drivers/pci.c Modified: trunk/openbios-devel/drivers/pci.c ============================================================================== --- trunk/openbios-devel/drivers/pci.c Fri Jan 8 13:09:44 2016 (r1375) +++ trunk/openbios-devel/drivers/pci.c Fri Jan 8 13:09:46 2016 (r1376) @@ -460,13 +460,18 @@ int ncells; ncells = 0; - /* first encode PCI configuration space */ - { - ncells += pci_encode_phys_addr(props + ncells, 0, CONFIGURATION_SPACE, + +#ifdef CONFIG_SPARC64 + /* While configuration space isn't mentioned in the IEEE-1275 PCI + bindings, it appears in the PCI host bridge ranges property in + real device trees. Hence we disable this range for all host + bridges except for SPARC, particularly as it causes Darwin/OS X + to incorrectly calculated PCI memory space ranges on PPC. */ + ncells += pci_encode_phys_addr(props + ncells, 0, CONFIGURATION_SPACE, config->dev, 0, 0); ncells += host_encode_phys_addr(props + ncells, arch->cfg_addr); ncells += pci_encode_size(props + ncells, arch->cfg_len); - } +#endif if (arch->io_base) { ncells += pci_encode_phys_addr(props + ncells, 0, IO_SPACE,

1 0