2009/5/14 Jacek Politowski <j...@jp.pl.eu.org>: > On Thu, May 14, 2009 at 09:47:26AM +0200, Michel Dänzer wrote: >>On Wed, 2009-05-13 at 13:13 +0200, Jacek Politowski wrote: > >>> I started git bisect between revisions >>> a6561f2ec673b38907f7181235386f32e60c32ba and >>> da021c36bbdf3bca31ee50ebe01cdb9495c09b36 >>> (...) (the same error for every tested build). > >>I think commit da021c36bbdf3bca31ee50ebe01cdb9495c09b36 fixed this, >>so >>something like > >>git show da021c36bbdf3bca31ee50ebe01cdb9495c09b36 | patch -p1 > > Ok, I've applied this patch and continued git bisect session. > > Driver builds correctly, so I was able to finish the search. > > Result follows: > $ git bisect visualize > commit b01e35168de40d192fd7a9ce6884b9c7419afbd4 > Author: Alex Deucher <alexdeuc...@gmail.com> > Date: Mon Nov 24 21:06:42 2008 -0500 > > Fix up posted logic > > noticed by benh on IRC > > > This is the first revision that shows symptoms from my initial > bug report.
Does this patch (against master) fix it? Alex
diff --git a/src/radeon_bios.c b/src/radeon_bios.c index 9907e09..ecf5403 100644 --- a/src/radeon_bios.c +++ b/src/radeon_bios.c @@ -328,6 +328,7 @@ radeon_card_posted(ScrnInfoPtr pScrn) unsigned char *RADEONMMIO = info->MMIO; uint32_t reg; + /* first check CRTCs */ if (IS_AVIVO_VARIANT) { reg = INREG(AVIVO_D1CRTC_CONTROL) | INREG(AVIVO_D2CRTC_CONTROL); if (reg & AVIVO_CRTC_EN) @@ -338,6 +339,15 @@ radeon_card_posted(ScrnInfoPtr pScrn) return TRUE; } + /* then check MEM_SIZE, in case something turned the crtcs off */ + if (info->ChipFamily >= CHIP_FAMILY_R600) + reg = INREG(R600_CONFIG_MEMSIZE); + else + reg = INREG(RADEON_CONFIG_MEMSIZE); + + if (reg) + return TRUE; + return FALSE; }