I have a Micron Millennium Transport laptop with Cirrus CL-GD7548
based video. It works fine in 16bpp using the Cirrus XFree86 4.2.0
driver. /dev/fb works fine at 8bpp, but didn't work at 16bpp. I
looked at the two drivers and made clgenfb.c do a few things the
XFree86 driver was doing and now 16bpp appears to work fine on that
laptop.
I don't subscribe to linux-fbdev-devel, but do read lkml via a
leisurely Usenet feed.
--Cliff Matthews After some people managed to scale the wall,
[email protected] there was a ban on the sale of rope and twine.
diff -Naur linux-2.4.20/drivers/video/clgenfb.c linux-2.4.20-clpatch/drivers/video/clgenfb.c
--- linux-2.4.20/drivers/video/clgenfb.c 2002-11-28 16:53:15.000000000 -0700
+++ linux-2.4.20-clpatch/drivers/video/clgenfb.c 2002-12-07 23:34:54.000000000 -0700
@@ -15,6 +15,9 @@
* Lars Hecking:
* Amiga updates and testing.
*
+ * Cliff Matthews <[email protected]>:
+ * 16bpp fix for CL-GD7548 (uses info from XFree86 4.2.0 source)
+ *
* Original clgenfb author: Frank Neumann
*
* Based on retz3fb.c and clgen.c:
@@ -403,6 +406,9 @@
#ifdef CONFIG_PCI
struct pci_dev *pdev;
+#define IS_7548(x) ((x)->pdev->device == PCI_DEVICE_ID_CIRRUS_7548)
+#else
+#define IS_7548(x) (FALSE)
#endif
};
@@ -970,7 +976,10 @@
DPRINTK ("desired pixclock: %ld kHz\n", freq);
- maxclock = clgen_board_info[fb_info->btype].maxclock;
+ if (IS_7548(fb_info))
+ maxclock = 80100;
+ else
+ maxclock = clgen_board_info[fb_info->btype].maxclock;
_par->multiplexing = 0;
/* If the frequency is greater than we can support, we might be able
@@ -1478,10 +1487,17 @@
case BT_ALPINE:
DPRINTK (" (for GD543x)\n");
- if (_par->HorizRes >= 1024)
- vga_wseq (fb_info->regs, CL_SEQR7, 0xa7);
- else
- vga_wseq (fb_info->regs, CL_SEQR7, 0xa3);
+ if (IS_7548(fb_info)) {
+ vga_wseq (fb_info->regs, CL_SEQR7,
+ (vga_rseq (fb_info->regs, CL_SEQR7) & 0xE0)
+ | 0x17);
+ WHDR (fb_info, 0xC1);
+ } else {
+ if (_par->HorizRes >= 1024)
+ vga_wseq (fb_info->regs, CL_SEQR7, 0xa7);
+ else
+ vga_wseq (fb_info->regs, CL_SEQR7, 0xa3);
+ }
clgen_set_mclk (fb_info, _par->mclk, _par->divMCLK);
break;
@@ -1594,6 +1610,11 @@
_par->var.bits_per_pixel);
}
+ if (IS_7548(fb_info)) {
+ vga_wseq (fb_info->regs, CL_SEQR2D,
+ vga_rseq (fb_info->regs, CL_SEQR2D) | 0xC0);
+ }
+
vga_wcrt (fb_info->regs, VGA_CRTC_OFFSET, offset & 0xff);
tmp = 0x22;
if (offset & 0x100)
diff -Naur linux-2.4.20/drivers/video/clgenfb.h linux-2.4.20-clpatch/drivers/video/clgenfb.h
--- linux-2.4.20/drivers/video/clgenfb.h 2000-01-06 11:23:46.000000000 -0700
+++ linux-2.4.20-clpatch/drivers/video/clgenfb.h 2002-12-07 20:50:43.000000000 -0700
@@ -60,6 +60,7 @@
#define CL_SEQR1D 0x1d /* VCLK2 Denominator and Post-Scalar Value */
#define CL_SEQR1E 0x1e /* VCLK3 Denominator and Post-Scalar Value */
#define CL_SEQR1F 0x1f /* BIOS ROM write enable and MCLK Select */
+#define CL_SEQR2D 0x2d /* unknown, snagged from XFree86 4.2.0 */
/*** CRT Controller Registers ***/
#define CL_CRT22 0x22 /* Graphics Data Latches ReadBack */
Adapted to 2.6.0 and applied. AT present I'm going threw my patch queue
to clean it out.
> I have a Micron Millennium Transport laptop with Cirrus CL-GD7548
> based video. It works fine in 16bpp using the Cirrus XFree86 4.2.0
> driver. /dev/fb works fine at 8bpp, but didn't work at 16bpp. I
> looked at the two drivers and made clgenfb.c do a few things the
> XFree86 driver was doing and now 16bpp appears to work fine on that
> laptop.
>
> I don't subscribe to linux-fbdev-devel, but do read lkml via a
> leisurely Usenet feed.
>
> --Cliff Matthews After some people managed to scale the wall,
> [email protected] there was a ban on the sale of rope and twine.
>
> diff -Naur linux-2.4.20/drivers/video/clgenfb.c linux-2.4.20-clpatch/drivers/video/clgenfb.c
> --- linux-2.4.20/drivers/video/clgenfb.c 2002-11-28 16:53:15.000000000 -0700
> +++ linux-2.4.20-clpatch/drivers/video/clgenfb.c 2002-12-07 23:34:54.000000000 -0700
> @@ -15,6 +15,9 @@
> * Lars Hecking:
> * Amiga updates and testing.
> *
> + * Cliff Matthews <[email protected]>:
> + * 16bpp fix for CL-GD7548 (uses info from XFree86 4.2.0 source)
> + *
> * Original clgenfb author: Frank Neumann
> *
> * Based on retz3fb.c and clgen.c:
> @@ -403,6 +406,9 @@
>
> #ifdef CONFIG_PCI
> struct pci_dev *pdev;
> +#define IS_7548(x) ((x)->pdev->device == PCI_DEVICE_ID_CIRRUS_7548)
> +#else
> +#define IS_7548(x) (FALSE)
> #endif
> };
>
> @@ -970,7 +976,10 @@
>
> DPRINTK ("desired pixclock: %ld kHz\n", freq);
>
> - maxclock = clgen_board_info[fb_info->btype].maxclock;
> + if (IS_7548(fb_info))
> + maxclock = 80100;
> + else
> + maxclock = clgen_board_info[fb_info->btype].maxclock;
> _par->multiplexing = 0;
>
> /* If the frequency is greater than we can support, we might be able
> @@ -1478,10 +1487,17 @@
>
> case BT_ALPINE:
> DPRINTK (" (for GD543x)\n");
> - if (_par->HorizRes >= 1024)
> - vga_wseq (fb_info->regs, CL_SEQR7, 0xa7);
> - else
> - vga_wseq (fb_info->regs, CL_SEQR7, 0xa3);
> + if (IS_7548(fb_info)) {
> + vga_wseq (fb_info->regs, CL_SEQR7,
> + (vga_rseq (fb_info->regs, CL_SEQR7) & 0xE0)
> + | 0x17);
> + WHDR (fb_info, 0xC1);
> + } else {
> + if (_par->HorizRes >= 1024)
> + vga_wseq (fb_info->regs, CL_SEQR7, 0xa7);
> + else
> + vga_wseq (fb_info->regs, CL_SEQR7, 0xa3);
> + }
> clgen_set_mclk (fb_info, _par->mclk, _par->divMCLK);
> break;
>
> @@ -1594,6 +1610,11 @@
> _par->var.bits_per_pixel);
> }
>
> + if (IS_7548(fb_info)) {
> + vga_wseq (fb_info->regs, CL_SEQR2D,
> + vga_rseq (fb_info->regs, CL_SEQR2D) | 0xC0);
> + }
> +
> vga_wcrt (fb_info->regs, VGA_CRTC_OFFSET, offset & 0xff);
> tmp = 0x22;
> if (offset & 0x100)
> diff -Naur linux-2.4.20/drivers/video/clgenfb.h linux-2.4.20-clpatch/drivers/video/clgenfb.h
> --- linux-2.4.20/drivers/video/clgenfb.h 2000-01-06 11:23:46.000000000 -0700
> +++ linux-2.4.20-clpatch/drivers/video/clgenfb.h 2002-12-07 20:50:43.000000000 -0700
> @@ -60,6 +60,7 @@
> #define CL_SEQR1D 0x1d /* VCLK2 Denominator and Post-Scalar Value */
> #define CL_SEQR1E 0x1e /* VCLK3 Denominator and Post-Scalar Value */
> #define CL_SEQR1F 0x1f /* BIOS ROM write enable and MCLK Select */
> +#define CL_SEQR2D 0x2d /* unknown, snagged from XFree86 4.2.0 */
>
> /*** CRT Controller Registers ***/
> #define CL_CRT22 0x22 /* Graphics Data Latches ReadBack */
>
>
> -------------------------------------------------------
> This sf.net email is sponsored by:ThinkGeek
> Welcome to geek heaven.
> http://thinkgeek.com/sf
> _______________________________________________
> Linux-fbdev-devel mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/linux-fbdev-devel
>