Gentoo Archives: gentoo-sparc

From: Ferris McCormick <fmccor@g.o>
To: Gentoo Sparc <gentoo-sparc@l.g.o>
Cc: x11@g.o, jbarnes@××××××××××××.org
Subject: [gentoo-sparc] [Fwd: [Bug 3914] ATI Mach64 locks up sparc post 6.8.99.3]
Date: Thu, 24 Aug 2006 18:51:00
Message-Id: 1156445384.12404.38.camel@liasis.inforead.com
1 Is anyone in a position to verify this? It's basically a proposed
2 one-line fix for the sparc/mach64 lockup problem with xorg-x11-7.1, and
3 if it works, it's better than remove-altix.
4
5 The actual proposed patch looks like
6 ========================
7 --- hw/xfree86/os-support/bus/linuxPci.c- 2006-08-24
8 18:32:52.000000000 +0000
9 +++ hw/xfree86/os-support/bus/linuxPci.c 2006-08-24
10 18:37:10.000000000 +0000
11 @@ -84,7 +84,11 @@
12 /* pciAddrBusToHost */ linuxPpcBusAddrToHostAddr,
13 #else
14 /* pciAddrHostToBus */ pciAddrNOOP,
15 +#ifdef __ia64__
16 /* pciAddrBusToHost */ linuxTransAddrBusToHost,
17 +#else
18 +/* pciAddrBusToHost */ pciAddrNOOP,
19 +#endif
20 #endif
21
22 /* pciControlBridge */ NULL,
23 ===========================
24 I have attached it as jparnes.patch, as well. I have verified that the
25 patch installs, and that linuxPci.c compiles with it applied, but I do
26 not have a mach64 system on which to test it.
27
28
29 -------- Forwarded Message --------
30 From: bugzilla-daemon@××××××××××××××××××××.org
31 To: fmccor@g.o
32 Subject: [Bug 3914] ATI Mach64 locks up sparc post 6.8.99.3
33 Date: Wed, 23 Aug 2006 09:20:20 -0700 (PDT)
34
35 Please do not reply to this email: if you want to comment on the bug, go to
36
37 the URL shown below and enter yourcomments there.
38
39 https://bugs.freedesktop.org/show_bug.cgi?id=3914
40
41
42
43
44
45 ------- Additional Comments From jbarnes@××××××××××××.org 2006-08-23 09:20 -------
46 The snippet below (from the revert patch) removes the new
47 linuxTransAddrBusToHost routine from the PCI mapping function table. This
48 routine is *supposed* to be generic, but it may well be broken on some sparc64
49 configurations. Can you try building a tree without the revert patch but with
50 that line protected #if defined(__ia64__) ... #endif instead of removed
51 entirely? That would tell us for sure if the xf86GetOSOffsetFromPCI routine
52 was the culprit (it probably is, it has quite a few shortcomings).
53
54 If that works, simply protecting that line with an #ifdef __ia64__ might be a
55 good short term fix...
56
57 Thanks,
58 Jesse
59
60 --- a/hw/xfree86/os-support/bus/linuxPci.c
61 +++ b/hw/xfree86/os-support/bus/linuxPci.c
62 @@ -63,7 +63,6 @@ #include "Pci.h"
63 static CARD32 linuxPciCfgRead(PCITAG tag, int off);
64 static void linuxPciCfgWrite(PCITAG, int off, CARD32 val);
65 static void linuxPciCfgSetBits(PCITAG tag, int off, CARD32 mask, CARD32 bits);
66 -static ADDRESS linuxTransAddrBusToHost(PCITAG tag, PciAddrType type, ADDRESS
67 addr);
68 #if defined(__powerpc__)
69 static ADDRESS linuxPpcBusAddrToHostAddr(PCITAG, PciAddrType, ADDRESS);
70 static ADDRESS linuxPpcHostAddrToBusAddr(PCITAG, PciAddrType, ADDRESS);
71 @@ -84,7 +83,7 @@ #if defined(__powerpc__)
72 /* pciAddrBusToHost */ linuxPpcBusAddrToHostAddr,
73 #else
74 /* pciAddrHostToBus */ pciAddrNOOP,
75 -/* pciAddrBusToHost */ linuxTransAddrBusToHost,
76 +/* pciAddrBusToHost */ pciAddrNOOP,
77 #endif
78
79 /* pciControlBridge */ NULL,
80
81
82 --
83 Configure bugmail: https://bugs.freedesktop.org/userprefs.cgi?tab=email
84
85 ------- You are receiving this mail because: -------
86 You are on the CC list for the bug, or are watching someone who is.
87
88 =================================================
89
90 Thanks,
91 Ferris
92 --
93 Ferris McCormick (P44646, MI) <fmccor@g.o>
94 Developer, Gentoo Linux (Devrel, Sparc)

Attachments

File name MIME type
jbarnes.patch text/x-patch
signature.asc application/pgp-signature