From: Tony Lindgren <tony@atomide.com>
To: Tuomas Kulve <tuomas.kulve@movial.fi>
Cc: linux-omap@vger.kernel.org
Subject: Re: Patch to enable omapfb scaling
Date: Wed, 26 Nov 2008 09:29:11 -0800 [thread overview]
Message-ID: <20081126172911.GN11997@atomide.com> (raw)
In-Reply-To: <490F12A5.1000900@movial.fi>
* Tuomas Kulve <tuomas.kulve@movial.fi> [081103 07:04]:
> Tuomas Kulve wrote:
> > Hi
> >
> > The attached patch enables omapfb scaling. Tested on omap3 based
> > beagleboard with 2.6.27-omap1 kernel from linux-omap kernel tree.
>
> The 0002 patch implements downscaling in omapfb. It seems to work on
> ES3.0 beagleboard and ES2.1 EVM but not on my rev B5 beagleboard.
>
> The patch still needs cleanup but it should work with e.g. X.Org's XV
> extension:
>
> http://gitweb.pingu.fi/?p=xf86-video-omapfb.git;a=summary
All, I'm not going to push any more omap video patches, those need to
be discussed on fbdev mailing list. So please send the patches against
the mainline kernel there and cc linux-omap mailing list.
Regards,
Tony
>
>
> --
> Tuomas Kulve, Movial Creative Technologies Inc.
> Porkkalankatu 13J, FI-00180 Helsinki
> Fax +358 9 8567 6401, Tel +358 9 8567 6400
> www.movial.com
> From 069ce763b1a7e936b7b2bdc288d94dd98e477d7e Mon Sep 17 00:00:00 2001
> From: Tuomas Kulve <tuomas.kulve@movial.com>
> Date: Mon, 3 Nov 2008 11:22:56 +0200
> Subject: [PATCH] Implemented video downsampling in omapfb.
>
> ---
> drivers/video/omap/dispc.c | 58 ++++++++++++++++++++++++++++++++++++++++----
> 1 files changed, 53 insertions(+), 5 deletions(-)
>
> diff --git a/drivers/video/omap/dispc.c b/drivers/video/omap/dispc.c
> index 68bc887..83f37bf 100644
> --- a/drivers/video/omap/dispc.c
> +++ b/drivers/video/omap/dispc.c
> @@ -545,6 +545,17 @@ static void write_firhv_reg(int plane, int reg, u32 value)
> dispc_write_reg(base + reg * 8, value);
> }
>
> +static void write_firv_reg(int plane, int reg, u32 value)
> +{
> + u32 base;
> +
> + if (plane == 1)
> + base = 0x1E0;
> + else
> + base = 0x1E0 + 0x20;
> + dispc_write_reg(base + reg * 4, value);
> +}
> +
> static void set_upsampling_coef_table(int plane)
> {
> const u32 coef[][2] = {
> @@ -565,6 +576,27 @@ static void set_upsampling_coef_table(int plane)
> }
> }
>
> +static void set_downsampling_coef_table(int plane)
> +{
> + const u32 coef[][3] = {
> + { 0x24382400, 0x24382400, 0x00000000 },
> + { 0x28371FFE, 0x28391F04, 0x000004FE },
> + { 0x2C361BFB, 0x2D381B08, 0x000008FB },
> + { 0x303516F9, 0x3237170C, 0x00000CF9 },
> + { 0x11343311, 0x123737F7, 0x0000F711 },
> + { 0x1635300C, 0x173732F9, 0x0000F90C },
> + { 0x1B362C08, 0x1B382DFB, 0x0000FB08 },
> + { 0x1F372804, 0x1F3928FE, 0x0000FE04 },
> + };
> + int i;
> +
> + for (i = 0; i < 8; i++) {
> + write_firh_reg(plane, i, coef[i][0]);
> + write_firhv_reg(plane, i, coef[i][1]);
> + write_firv_reg(plane, i, coef[i][2]);
> + }
> +}
> +
> static int omap_dispc_set_scale(int plane,
> int orig_width, int orig_height,
> int out_width, int out_height)
> @@ -592,16 +624,32 @@ static int omap_dispc_set_scale(int plane,
> if (orig_height > out_height ||
> orig_width * 8 < out_width ||
> orig_height * 8 < out_height) {
> + dev_dbg(dispc.fbdev->dev,
> + "Max upsampling is 8x, "
> + "tried: %dx%d -> %dx%d\n",
> + orig_width, orig_height,
> + out_width, out_height);
> enable_lcd_clocks(0);
> return -EINVAL;
> }
> set_upsampling_coef_table(plane);
> } else if (orig_width > out_width) {
> - /* Downsampling not yet supported
> - */
> -
> - enable_lcd_clocks(0);
> - return -EINVAL;
> + /*
> + * Downsampling.
> + * Currently you can only scale both dimensions in one way.
> + */
> + if (orig_height < out_height ||
> + orig_width > out_width * 4 ||
> + orig_height > out_height * 4) {
> + dev_dbg(dispc.fbdev->dev,
> + "Max downsampling is 4x, "
> + "tried: %dx%d -> %dx%d\n",
> + orig_width, orig_height,
> + out_width, out_height);
> + enable_lcd_clocks(0);
> + return -EINVAL;
> + }
> + set_downsampling_coef_table(plane);
> }
> if (!orig_width || orig_width == out_width)
> fir_hinc = 0;
> --
> 1.5.6.5
>
> From fd41a240a99de4c1408097bb39fa89cec4b7868d Mon Sep 17 00:00:00 2001
> From: Tuomas Kulve <tuomas.kulve@movial.com>
> Date: Fri, 31 Oct 2008 20:05:26 +0200
> Subject: [PATCH] Calculate fir_*inc using resolution minus one in omapfb scaling.
>
> ---
> drivers/video/omap/dispc.c | 4 ++--
> 1 files changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/video/omap/dispc.c b/drivers/video/omap/dispc.c
> index 83f37bf..7ac4a86 100644
> --- a/drivers/video/omap/dispc.c
> +++ b/drivers/video/omap/dispc.c
> @@ -654,11 +654,11 @@ static int omap_dispc_set_scale(int plane,
> if (!orig_width || orig_width == out_width)
> fir_hinc = 0;
> else
> - fir_hinc = 1024 * orig_width / out_width;
> + fir_hinc = 1024 * (orig_width - 1) / (out_width - 1);
> if (!orig_height || orig_height == out_height)
> fir_vinc = 0;
> else
> - fir_vinc = 1024 * orig_height / out_height;
> + fir_vinc = 1024 * (orig_height - 1) / (out_height - 1);
> dispc.fir_hinc[plane] = fir_hinc;
> dispc.fir_vinc[plane] = fir_vinc;
>
> --
> 1.5.6.5
>
prev parent reply other threads:[~2008-11-26 17:29 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2008-10-31 10:44 Patch to enable omapfb scaling Tuomas Kulve
2008-11-03 15:03 ` Tuomas Kulve
2008-11-26 17:29 ` Tony Lindgren [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20081126172911.GN11997@atomide.com \
--to=tony@atomide.com \
--cc=linux-omap@vger.kernel.org \
--cc=tuomas.kulve@movial.fi \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox