On Mon, Sep 8, 2008 at 17:59:11 -0400, Anthony DeRobertis wrote:
> 000:<:0014: 32: Request(98): QueryExtension name='XFree86-VidModeExtension'
> 000:>:0x0014:32: Reply to QueryExtension: present=true(0x01) major-opcode=134
> first-event=0 first-error=130
> 000:<:0015: 4: XFree86-VidModeExtension-Request(134,0):
> XF86VidModeQueryVersion
> 000:>:0x0015:32: Reply to XF86VidModeQueryVersion: major-version=2
> minor-version=2
> 000:<:0016: 8: XFree86-VidModeExtension-Request(134,14):
> XF86VidModeSetClientVersion major=2 minor=2
> 000:<:0017: 8: XFree86-VidModeExtension-Request(134,1):
> XF86VidModeGetModeLine screen=0
> [EMAIL PROTECTED]:~$
>
Thanks for all the debugging! Can you try the attached patch (against
the xorg-server source)?
Cheers,
Julien
>From c8af7ce35a900ac9b898f51c1b95dabad3ba1d76 Mon Sep 17 00:00:00 2001
From: Paulo Cesar Pereira de Andrade <[EMAIL PROTECTED]>
Date: Fri, 23 May 2008 13:50:39 -0300
Subject: [PATCH] Correct a NULL pointer deference
The problem happens if Monitor/Card combo doesn't provide EDID info,
and the XFree86-VidModeExtension extension is used.
Signed-off-by: Peter Hutterer <[EMAIL PROTECTED]>
---
hw/xfree86/common/xf86VidMode.c | 10 +++++++---
1 files changed, 7 insertions(+), 3 deletions(-)
diff --git a/hw/xfree86/common/xf86VidMode.c b/hw/xfree86/common/xf86VidMode.c
index 38d605c..9260bb5 100644
--- a/hw/xfree86/common/xf86VidMode.c
+++ b/hw/xfree86/common/xf86VidMode.c
@@ -150,10 +150,14 @@ VidModeGetCurrentModeline(int scrnIndex, pointer *mode, int *dotClock)
return FALSE;
pScrn = xf86Screens[scrnIndex];
- *mode = (pointer)(pScrn->currentMode);
- *dotClock = pScrn->currentMode->Clock;
- return TRUE;
+ if (pScrn->currentMode) {
+ *mode = (pointer)(pScrn->currentMode);
+ *dotClock = pScrn->currentMode->Clock;
+
+ return TRUE;
+ }
+ return FALSE;
}
_X_EXPORT int
--
1.5.6.5
