* [PATCH] i.MX23/28 framebuffer driver
@ 2011-02-09 13:20 Sascha Hauer
2011-02-09 13:20 ` Sascha Hauer
` (2 more replies)
0 siblings, 3 replies; 65+ messages in thread
From: Sascha Hauer @ 2011-02-09 13:20 UTC (permalink / raw)
To: linux-arm-kernel
Hi all,
The following patches add framebuffer support for Freescale i.MX23/28
processors.
Sascha Hauer (2):
video: Add i.MX23/28 framebuffer driver
ARM i.MX23/28: Add framebuffer device support
arch/arm/mach-mxs/clock-mx23.c | 2 +-
arch/arm/mach-mxs/clock-mx28.c | 1 +
arch/arm/mach-mxs/devices-mx23.h | 4 +
arch/arm/mach-mxs/devices-mx28.h | 4 +
arch/arm/mach-mxs/devices/Kconfig | 4 +
arch/arm/mach-mxs/devices/Makefile | 1 +
arch/arm/mach-mxs/devices/platform-mxsfb.c | 46 ++
arch/arm/mach-mxs/include/mach/fb.h | 48 ++
drivers/video/Kconfig | 9 +
drivers/video/Makefile | 1 +
drivers/video/mxsfb.c | 914 ++++++++++++++++++++++++++++
11 files changed, 1033 insertions(+), 1 deletions(-)
create mode 100644 arch/arm/mach-mxs/devices/platform-mxsfb.c
create mode 100644 arch/arm/mach-mxs/include/mach/fb.h
create mode 100644 drivers/video/mxsfb.c
^ permalink raw reply [flat|nested] 65+ messages in thread
* [PATCH 1/2] video: Add i.MX23/28 framebuffer driver
2011-02-09 13:20 [PATCH] i.MX23/28 framebuffer driver Sascha Hauer
@ 2011-02-09 13:20 ` Sascha Hauer
2011-02-09 13:20 ` [PATCH 2/2] ARM i.MX23/28: Add framebuffer device support Sascha Hauer
2011-02-09 14:47 ` [PATCH] i.MX23/28 framebuffer driver Arnd Bergmann
2 siblings, 0 replies; 65+ messages in thread
From: Sascha Hauer @ 2011-02-09 13:20 UTC (permalink / raw)
To: linux-arm-kernel
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Cc: Paul Mundt <lethal@linux-sh.org>
Cc: linux-fbdev@vger.kernel.org
Cc: Shawn Guo <shawn.guo@freescale.com>
---
arch/arm/mach-mxs/include/mach/fb.h | 48 ++
drivers/video/Kconfig | 9 +
drivers/video/Makefile | 1 +
drivers/video/mxsfb.c | 914 +++++++++++++++++++++++++++++++++++
4 files changed, 972 insertions(+), 0 deletions(-)
create mode 100644 arch/arm/mach-mxs/include/mach/fb.h
create mode 100644 drivers/video/mxsfb.c
diff --git a/arch/arm/mach-mxs/include/mach/fb.h b/arch/arm/mach-mxs/include/mach/fb.h
new file mode 100644
index 0000000..923f397
--- /dev/null
+++ b/arch/arm/mach-mxs/include/mach/fb.h
@@ -0,0 +1,48 @@
+/*
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation; either version 2
+ * of the License, or (at your option) any later version.
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
+ * MA 02110-1301, USA.
+ */
+
+#ifndef __MACH_FB_H
+#define __MACH_FB_H
+
+#include <linux/fb.h>
+
+#define STMLCDIF_8BIT 1 /** pixel data bus to the display is of 8 bit width */
+#define STMLCDIF_16BIT 0 /** pixel data bus to the display is of 16 bit width */
+#define STMLCDIF_18BIT 2 /** pixel data bus to the display is of 18 bit width */
+#define STMLCDIF_24BIT 3 /** pixel data bus to the display is of 24 bit width */
+
+#define FB_SYNC_DATA_ENABLE_HIGH_ACT 64
+
+struct mxsfb_platform_data {
+ struct fb_videomode *mode_list;
+ unsigned mode_count;
+
+ unsigned default_bpp;
+
+ unsigned dotclk_delay; /* refer manual HW_LCDIF_VDCTRL4 register */
+ unsigned ld_intf_width; /* refer STMLCDIF_* macros */
+
+ unsigned fb_size; /* Size of the video memory. If zero a
+ * default will be used
+ */
+ unsigned long fb_phys; /* physical address for the video memory. If
+ * zero the framebuffer memory will be dynamically
+ * allocated. If specified,fb_size must also be specified.
+ * fb_phys must be unused by Linux.
+ */
+};
+
+#endif /* __MACH_FB_H */
diff --git a/drivers/video/Kconfig b/drivers/video/Kconfig
index 6bafb51b..e0ea23f 100644
--- a/drivers/video/Kconfig
+++ b/drivers/video/Kconfig
@@ -2365,6 +2365,15 @@ config FB_JZ4740
help
Framebuffer support for the JZ4740 SoC.
+config FB_MXS
+ tristate "MXS LCD framebuffer support"
+ depends on FB && ARCH_MXS
+ select FB_CFB_FILLRECT
+ select FB_CFB_COPYAREA
+ select FB_CFB_IMAGEBLIT
+ help
+ Framebuffer support for the MXS SoC.
+
source "drivers/video/omap/Kconfig"
source "drivers/video/omap2/Kconfig"
diff --git a/drivers/video/Makefile b/drivers/video/Makefile
index 8c8fabd..9a096ae 100644
--- a/drivers/video/Makefile
+++ b/drivers/video/Makefile
@@ -153,6 +153,7 @@ obj-$(CONFIG_FB_BFIN_T350MCQB) += bfin-t350mcqb-fb.o
obj-$(CONFIG_FB_BFIN_7393) += bfin_adv7393fb.o
obj-$(CONFIG_FB_MX3) += mx3fb.o
obj-$(CONFIG_FB_DA8XX) += da8xx-fb.o
+obj-$(CONFIG_FB_MXS) += mxsfb.o
# the test framebuffer is last
obj-$(CONFIG_FB_VIRTUAL) += vfb.o
diff --git a/drivers/video/mxsfb.c b/drivers/video/mxsfb.c
new file mode 100644
index 0000000..92a5804
--- /dev/null
+++ b/drivers/video/mxsfb.c
@@ -0,0 +1,914 @@
+/*
+ * Copyright (C) 2010 Juergen Beisert, Pengutronix
+ *
+ * This code is based on:
+ * Author: Vitaly Wool <vital@embeddedalley.com>
+ *
+ * Copyright 2008-2009 Freescale Semiconductor, Inc. All Rights Reserved.
+ * Copyright 2008 Embedded Alley Solutions, Inc All Rights Reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation; either version 2
+ * of the License, or (at your option) any later version.
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ */
+
+#define DRIVER_NAME "mxsfb"
+
+/**
+ * @file
+ * @brief LCDIF driver for i.MX23 and i.MX28
+ *
+ * The LCDIF support four modes of operation
+ * - MPU interface (to drive smart displays) -> not supported yet
+ * - VSYNC interface (like MPU interface plus Vsync) -> not supported yet
+ * - Dotclock interface (to drive LC displays with RGB data and sync signals)
+ * - DVI (to drive ITU-R BT656) -> not supported yet
+ *
+ * This driver depends on a correct setup of the pins used for this purpose
+ * (platform specific).
+ *
+ * For the developer: Don't forget to set the data bus width to the display
+ * in the imx_fb_videomode structure. You will else end up with ugly colours.
+ * If you fight against jitter you can vary the clock delay. This is a feature
+ * of the i.MX28 and you can vary it between 2 ns ... 8 ns in 2 ns steps. Give
+ * the required value in the imx_fb_videomode structure.
+ */
+
+#include <linux/kernel.h>
+#include <linux/platform_device.h>
+#include <linux/clk.h>
+#include <linux/dma-mapping.h>
+#include <linux/io.h>
+#include <mach/fb.h>
+#include <mach/hardware.h>
+#include <mach/mxs.h>
+
+#define CTRL 0x00
+#define CTRL1 0x10
+#define MX28_CTRL2 0x20
+#define MX23_TRANSFER_COUNT 0x20
+#define MX28_TRANSFER_COUNT 0x30
+#define MX28_CUR_BUF 0x40
+#define MX28_NEXT_BUF 0x50
+#define MX23_CUR_BUF 0x30
+#define MX23_NEXT_BUF 0x40
+#define TIMING 0x60
+#define VDCTRL0 0x70
+#define VDCTRL1 0x80
+#define VDCTRL2 0x90
+#define VDCTRL3 0xa0
+#define VDCTRL4 0xb0
+#define DVICTRL0 0xc0
+#define DVICTRL1 0xd0
+#define DVICTRL2 0xe0
+#define DVICTRL3 0xf0
+#define DVICTRL4 0x100
+#define MX28_DATA 0x180
+#define MX23_DATA 0x1b0
+#define MX28_DEBUG0 0x1d0
+#define MX23_DEBUG0 0x1f0
+
+#define CTRL_SFTRST (1 << 31)
+#define CTRL_CLKGATE (1 << 30)
+#define CTRL_BYPASS_COUNT (1 << 19)
+#define CTRL_VSYNC_MODE (1 << 18)
+#define CTRL_DOTCLK_MODE (1 << 17)
+#define CTRL_DATA_SELECT (1 << 16)
+#define CTRL_SET_BUS_WIDTH(x) (((x) & 0x3) << 10)
+#define CTRL_GET_BUS_WIDTH(x) (((x) >> 10) & 0x3)
+#define CTRL_SET_WORD_LENGTH(x) (((x) & 0x3) << 8)
+#define CTRL_GET_WORD_LENGTH(x) (((x) >> 8) & 0x3)
+#define CTRL_MASTER (1 << 5)
+#define CTRL_DF16 (1 << 3)
+#define CTRL_DF18 (1 << 2)
+#define CTRL_DF24 (1 << 1)
+#define CTRL_RUN (1 << 0)
+
+#define CTRL1_FIFO_CLEAR (1 << 21)
+#define CTRL1_SET_BYTE_PACKAGING(x) (((x) & 0xf) << 16)
+#define CTRL1_GET_BYTE_PACKAGING(x) (((x) >> 16) & 0xf)
+
+#define TRANSFER_COUNT_SET_VCOUNT(x) (((x) & 0xffff) << 16)
+#define TRANSFER_COUNT_GET_VCOUNT(x) (((x) >> 16) & 0xffff)
+#define TRANSFER_COUNT_SET_HCOUNT(x) ((x) & 0xffff)
+#define TRANSFER_COUNT_GET_HCOUNT(x) ((x) & 0xffff)
+
+
+#define VDCTRL0_ENABLE_PRESENT (1 << 28)
+#define VDCTRL0_VSYNC_ACT_HIGH (1 << 27)
+#define VDCTRL0_HSYNC_ACT_HIGH (1 << 26)
+#define VDCTRL0_DOTCLK_ACTIVE_HIGH (1 << 25)
+#define VDCTRL0_POL_ACTIVE_HIGH (1 << 24)
+#define VDCTRL0_VSYNC_PERIOD_UNIT (1 << 21)
+#define VDCTRL0_VSYNC_PULSE_WIDTH_UNIT (1 << 20)
+#define VDCTRL0_HALF_LINE (1 << 19)
+#define VDCTRL0_HALF_LINE_MODE (1 << 18)
+#define VDCTRL0_SET_VSYNC_PULSE_WIDTH(x) ((x) & 0x3ffff)
+#define VDCTRL0_GET_VSYNC_PULSE_WIDTH(x) ((x) & 0x3ffff)
+
+#define VDCTRL2_SET_HSYNC_PERIOD(x) ((x) & 0x3ffff)
+#define VDCTRL2_GET_HSYNC_PERIOD(x) ((x) & 0x3ffff)
+
+#define VDCTRL3_MUX_SYNC_SIGNALS (1 << 29)
+#define VDCTRL3_VSYNC_ONLY (1 << 28)
+#define SET_HOR_WAIT_CNT(x) (((x) & 0xfff) << 16)
+#define GET_HOR_WAIT_CNT(x) (((x) >> 16) & 0xfff)
+#define SET_VERT_WAIT_CNT(x) ((x) & 0xffff)
+#define GET_VERT_WAIT_CNT(x) ((x) & 0xffff)
+
+#define VDCTRL4_SET_DOTCLK_DLY(x) (((x) & 0x7) << 29) /* i.MX28 only */
+#define VDCTRL4_GET_DOTCLK_DLY(x) (((x) >> 29) & 0x7) /* i.MX28 only */
+#define VDCTRL4_SYNC_SIGNALS_ON (1 << 18)
+#define SET_DOTCLK_H_VALID_DATA_CNT(x) ((x) & 0x3ffff)
+
+#define DEBUG0_HSYNC (1 < 26)
+#define DEBUG0_VSYNC (1 < 25)
+
+#define MIN_XRES 120
+#define MIN_YRES 120
+
+#define RED 0
+#define GREEN 1
+#define BLUE 2
+#define TRANSP 3
+
+enum mxsfb_devtype {
+ MXSFB_MX23,
+ MXSFB_MX28,
+};
+
+/* CPU dependent register offsets */
+struct mxsfb_devdata {
+ unsigned transfer_count;
+ unsigned cur_buf;
+ unsigned next_buf;
+ unsigned debug0;
+ unsigned hs_wdth_mask;
+ unsigned hs_wdth_shift;
+};
+
+struct mxsfb_info {
+ struct fb_info fb_info;
+ struct platform_device *pdev;
+ struct clk *clk;
+ void __iomem *base; /* registers */
+ unsigned allocated_size;
+ int enabled;
+ unsigned ld_intf_width;
+ unsigned dotclk_delay;
+ const struct mxsfb_devdata *devdata;
+ int mapped;
+};
+
+static const struct mxsfb_devdata mxsfb_devdata[] = {
+ [MXSFB_MX23] = {
+ .transfer_count = MX23_TRANSFER_COUNT,
+ .cur_buf = MX23_CUR_BUF,
+ .next_buf = MX23_NEXT_BUF,
+ .debug0 = MX23_DEBUG0,
+ .hs_wdth_mask = 0xff,
+ .hs_wdth_shift = 24,
+ },
+ [MXSFB_MX28] = {
+ .transfer_count = MX28_TRANSFER_COUNT,
+ .cur_buf = MX28_CUR_BUF,
+ .next_buf = MX28_NEXT_BUF,
+ .debug0 = MX28_DEBUG0,
+ .hs_wdth_mask = 0x3fff,
+ .hs_wdth_shift = 18,
+ },
+};
+
+#define to_imxfb_host(x) (container_of(x, struct mxsfb_info, fb_info))
+
+/* mask and shift depends on architecture */
+static inline u32 set_hsync_pulse_width(struct mxsfb_info *host, unsigned val)
+{
+ return (val & host->devdata->hs_wdth_mask) <<
+ host->devdata->hs_wdth_shift;
+}
+
+static inline u32 get_hsync_pulse_width(struct mxsfb_info *host, unsigned val)
+{
+ return (val >> host->devdata->hs_wdth_shift) &
+ host->devdata->hs_wdth_mask;
+}
+
+static const struct fb_bitfield def_rgb565[] = {
+ [RED] = {
+ .offset = 11,
+ .length = 5,
+ },
+ [GREEN] = {
+ .offset = 5,
+ .length = 6,
+ },
+ [BLUE] = {
+ .offset = 0,
+ .length = 5,
+ },
+ [TRANSP] = { /* no support for transparency */
+ .length = 0,
+ }
+};
+
+static const struct fb_bitfield def_rgb666[] = {
+ [RED] = {
+ .offset = 16,
+ .length = 6,
+ },
+ [GREEN] = {
+ .offset = 8,
+ .length = 6,
+ },
+ [BLUE] = {
+ .offset = 0,
+ .length = 6,
+ },
+ [TRANSP] = { /* no support for transparency */
+ .length = 0,
+ }
+};
+
+static const struct fb_bitfield def_rgb888[] = {
+ [RED] = {
+ .offset = 16,
+ .length = 8,
+ },
+ [GREEN] = {
+ .offset = 8,
+ .length = 8,
+ },
+ [BLUE] = {
+ .offset = 0,
+ .length = 8,
+ },
+ [TRANSP] = { /* no support for transparency */
+ .length = 0,
+ }
+};
+
+static inline unsigned chan_to_field(unsigned chan, struct fb_bitfield *bf)
+{
+ chan &= 0xffff;
+ chan >>= 16 - bf->length;
+ return chan << bf->offset;
+}
+
+static int mxsfb_check_var(struct fb_var_screeninfo *var,
+ struct fb_info *fb_info)
+{
+ struct mxsfb_info *host = to_imxfb_host(fb_info);
+ const struct fb_bitfield *rgb = NULL;
+
+ if (var->xres < MIN_XRES)
+ var->xres = MIN_XRES;
+ if (var->yres < MIN_YRES)
+ var->yres = MIN_YRES;
+
+ var->xres_virtual = var->xres;
+
+ var->yres_virtual = var->yres;
+
+ switch (var->bits_per_pixel) {
+ case 16:
+ /* always expect RGB 565 */
+ rgb = def_rgb565;
+ break;
+ case 32:
+ switch (host->ld_intf_width) {
+ case STMLCDIF_8BIT:
+ pr_debug("Unsupported LCD bus width mapping\n");
+ break;
+ case STMLCDIF_16BIT:
+ case STMLCDIF_18BIT:
+ /* 24 bit to 18 bit mapping */
+ rgb = def_rgb666;
+ break;
+ case STMLCDIF_24BIT:
+ /* real 24 bit */
+ rgb = def_rgb888;
+ break;
+ }
+ break;
+ default:
+ pr_debug("Unsupported colour depth: %u\n", var->bits_per_pixel);
+ return -EINVAL;
+ }
+
+ /*
+ * Copy the RGB parameters for this display
+ * from the machine specific parameters.
+ */
+ var->red = rgb[RED];
+ var->green = rgb[GREEN];
+ var->blue = rgb[BLUE];
+ var->transp = rgb[TRANSP];
+
+ return 0;
+}
+
+static void mxsfb_enable_controller(struct fb_info *fb_info)
+{
+ struct mxsfb_info *host = to_imxfb_host(fb_info);
+ u32 reg;
+
+ dev_dbg(&host->pdev->dev, "%s\n", __func__);
+
+ clk_enable(host->clk);
+ clk_set_rate(host->clk, PICOS2KHZ(fb_info->var.pixclock) * 1000U);
+
+ /* if it was disabled, re-enable the mode again */
+ reg = readl(host->base + CTRL);
+ reg |= CTRL_DOTCLK_MODE;
+ writel(reg, host->base + CTRL);
+
+ /* enable the SYNC signals first, then the DMA engine */
+ reg = readl(host->base + VDCTRL4);
+ reg |= VDCTRL4_SYNC_SIGNALS_ON;
+ writel(reg, host->base + VDCTRL4);
+
+ reg = readl(host->base + CTRL);
+ reg |= CTRL_RUN;
+ writel(reg, host->base + CTRL);
+
+ host->enabled = 1;
+}
+
+static void mxsfb_disable_controller(struct fb_info *fb_info)
+{
+ struct mxsfb_info *host = to_imxfb_host(fb_info);
+ unsigned loop;
+ u32 reg;
+
+ dev_dbg(&host->pdev->dev, "%s\n", __func__);
+
+ /*
+ * Even if we disable the controller here, it will still continue
+ * until its FIFOs are running out of data
+ */
+ reg = readl(host->base + CTRL);
+ reg &= ~CTRL_DOTCLK_MODE;
+ writel(reg, host->base + CTRL);
+
+ loop = 1000;
+ while (loop) {
+ reg = readl(host->base + CTRL);
+ if (!(reg & CTRL_RUN))
+ break;
+ loop--;
+ }
+
+ reg = readl(host->base + VDCTRL4);
+ reg &= ~VDCTRL4_SYNC_SIGNALS_ON;
+ writel(reg, host->base + VDCTRL4);
+
+ clk_disable(host->clk);
+
+ host->enabled = 0;
+}
+
+static int mxsfb_set_par(struct fb_info *fb_info)
+{
+ struct mxsfb_info *host = to_imxfb_host(fb_info);
+ u32 ctrl, vdctrl0, vdctrl4;
+ int line_size, fb_size;
+ int reenable = 0;
+
+ line_size = fb_info->var.xres * (fb_info->var.bits_per_pixel >> 3);
+ fb_size = fb_info->var.yres_virtual * line_size;
+
+ if (fb_size > fb_info->fix.smem_len)
+ return -ENOMEM;
+
+ fb_info->fix.line_length = line_size;
+
+ /*
+ * It seems, you can't re-program the controller if it is still running.
+ * This may lead into shifted pictures (FIFO issue?).
+ * So, first stop the controller and drain its FIFOs
+ */
+ if (host->enabled) {
+ reenable = 1;
+ mxsfb_disable_controller(fb_info);
+ }
+
+ /* clear the FIFOs */
+ writel(CTRL1_FIFO_CLEAR, host->base + CTRL1 + 4);
+
+ ctrl = CTRL_BYPASS_COUNT | CTRL_MASTER |
+ CTRL_SET_BUS_WIDTH(host->ld_intf_width);;
+
+ switch (fb_info->var.bits_per_pixel) {
+ case 16:
+ dev_dbg(&host->pdev->dev, "Setting up RGB565 mode\n");
+ ctrl |= CTRL_SET_WORD_LENGTH(0);
+ writel(CTRL1_SET_BYTE_PACKAGING(0xf), host->base + CTRL1);
+ break;
+ case 32:
+ dev_dbg(&host->pdev->dev, "Setting up RGB888/666 mode\n");
+ ctrl |= CTRL_SET_WORD_LENGTH(3);
+ switch (host->ld_intf_width) {
+ case STMLCDIF_8BIT:
+ dev_dbg(&host->pdev->dev,
+ "Unsupported LCD bus width mapping\n");
+ return -EINVAL;
+ case STMLCDIF_16BIT:
+ case STMLCDIF_18BIT:
+ /* 24 bit to 18 bit mapping */
+ ctrl |= CTRL_DF24; /* ignore the upper 2 bits in
+ * each colour component
+ */
+ break;
+ case STMLCDIF_24BIT:
+ /* real 24 bit */
+ break;
+ }
+ /* do not use packed pixels = one pixel per word instead */
+ writel(CTRL1_SET_BYTE_PACKAGING(0x7), host->base + CTRL1);
+ break;
+ default:
+ dev_dbg(&host->pdev->dev, "Unhandled color depth of %u\n",
+ fb_info->var.bits_per_pixel);
+ return -EINVAL;
+ }
+
+ writel(ctrl, host->base + CTRL);
+
+ writel(TRANSFER_COUNT_SET_VCOUNT(fb_info->var.yres) |
+ TRANSFER_COUNT_SET_HCOUNT(fb_info->var.xres),
+ host->base + host->devdata->transfer_count);
+
+ vdctrl0 = VDCTRL0_ENABLE_PRESENT | /* always in DOTCLOCK mode */
+ VDCTRL0_VSYNC_PERIOD_UNIT |
+ VDCTRL0_VSYNC_PULSE_WIDTH_UNIT |
+ VDCTRL0_SET_VSYNC_PULSE_WIDTH(fb_info->var.vsync_len);
+ if (fb_info->var.sync & FB_SYNC_HOR_HIGH_ACT)
+ vdctrl0 |= VDCTRL0_HSYNC_ACT_HIGH;
+ if (fb_info->var.sync & FB_SYNC_VERT_HIGH_ACT)
+ vdctrl0 |= VDCTRL0_VSYNC_ACT_HIGH;
+ if (fb_info->var.sync & FB_SYNC_DATA_ENABLE_HIGH_ACT)
+ vdctrl0 |= VDCTRL0_POL_ACTIVE_HIGH;
+
+ writel(vdctrl0, host->base + VDCTRL0);
+
+ /* frame length in lines */
+ writel(fb_info->var.upper_margin + fb_info->var.vsync_len +
+ fb_info->var.lower_margin + fb_info->var.yres,
+ host->base + VDCTRL1);
+
+ /* line length in units of clocks or pixels */
+ writel(set_hsync_pulse_width(host, fb_info->var.hsync_len) |
+ VDCTRL2_SET_HSYNC_PERIOD(fb_info->var.left_margin +
+ fb_info->var.hsync_len + fb_info->var.right_margin +
+ fb_info->var.xres),
+ host->base + VDCTRL2);
+
+ writel(SET_HOR_WAIT_CNT(fb_info->var.left_margin +
+ fb_info->var.hsync_len) |
+ SET_VERT_WAIT_CNT(fb_info->var.upper_margin +
+ fb_info->var.vsync_len),
+ host->base + VDCTRL3);
+
+ vdctrl4 = SET_DOTCLK_H_VALID_DATA_CNT(fb_info->var.xres);
+ if (cpu_is_mx28())
+ vdctrl4 |= VDCTRL4_SET_DOTCLK_DLY(host->dotclk_delay);
+ writel(vdctrl4, host->base + VDCTRL4);
+
+ writel(fb_info->fix.smem_start +
+ fb_info->fix.line_length * fb_info->var.yoffset,
+ host->base + host->devdata->next_buf);
+
+ if (reenable)
+ mxsfb_enable_controller(fb_info);
+
+ return 0;
+}
+
+static int mxsfb_setcolreg(u_int regno, u_int red, u_int green, u_int blue,
+ u_int transp, struct fb_info *fb_info)
+{
+ unsigned int val;
+ int ret = -EINVAL;
+
+ /*
+ * If greyscale is true, then we convert the RGB value
+ * to greyscale no matter what visual we are using.
+ */
+ if (fb_info->var.grayscale)
+ red = green = blue = (19595 * red + 38470 * green +
+ 7471 * blue) >> 16;
+
+ switch (fb_info->fix.visual) {
+ case FB_VISUAL_TRUECOLOR:
+ /*
+ * 12 or 16-bit True Colour. We encode the RGB value
+ * according to the RGB bitfield information.
+ */
+ if (regno < 16) {
+ u32 *pal = fb_info->pseudo_palette;
+
+ val = chan_to_field(red, &fb_info->var.red);
+ val |= chan_to_field(green, &fb_info->var.green);
+ val |= chan_to_field(blue, &fb_info->var.blue);
+
+ pal[regno] = val;
+ ret = 0;
+ }
+ break;
+
+ case FB_VISUAL_STATIC_PSEUDOCOLOR:
+ case FB_VISUAL_PSEUDOCOLOR:
+ break;
+ }
+
+ return ret;
+}
+
+static int mxsfb_blank(int blank, struct fb_info *fb_info)
+{
+ struct mxsfb_info *host = to_imxfb_host(fb_info);
+
+ switch (blank) {
+ case FB_BLANK_POWERDOWN:
+ case FB_BLANK_VSYNC_SUSPEND:
+ case FB_BLANK_HSYNC_SUSPEND:
+ case FB_BLANK_NORMAL:
+ if (host->enabled)
+ mxsfb_disable_controller(fb_info);
+ break;
+
+ case FB_BLANK_UNBLANK:
+ if (!host->enabled)
+ mxsfb_enable_controller(fb_info);
+ break;
+ }
+ return 0;
+}
+
+static int mxsfb_pan_display(struct fb_var_screeninfo *var,
+ struct fb_info *fb_info)
+{
+ struct mxsfb_info *host = to_imxfb_host(fb_info);
+ unsigned offset;
+
+ if (var->xoffset != 0)
+ return -EINVAL;
+
+ offset = fb_info->fix.line_length * var->yoffset;
+
+ /* update on next VSYNC */
+ writel(fb_info->fix.smem_start + offset,
+ host->base + host->devdata->next_buf);
+
+ return 0;
+}
+
+static struct fb_ops mxsfb_ops = {
+ .owner = THIS_MODULE,
+ .fb_check_var = mxsfb_check_var,
+ .fb_set_par = mxsfb_set_par,
+ .fb_setcolreg = mxsfb_setcolreg,
+ .fb_blank = mxsfb_blank,
+ .fb_pan_display = mxsfb_pan_display,
+ .fb_fillrect = cfb_fillrect,
+ .fb_copyarea = cfb_copyarea,
+ .fb_imageblit = cfb_imageblit,
+};
+
+static int __devinit mxsfb_restore_mode(struct mxsfb_info *host)
+{
+ struct fb_info *fb_info = &host->fb_info;
+ unsigned line_count;
+ unsigned period;
+ unsigned long pa, fbsize;
+ int bits_per_pixel, ofs;
+ u32 transfer_count, vdctrl0, vdctrl2, vdctrl3, vdctrl4, ctrl;
+ struct fb_videomode vmode;
+
+ /* Only restore the mode when the controller is running */
+ ctrl = readl(host->base + CTRL);
+ if (!(ctrl & CTRL_RUN))
+ return -EINVAL;
+
+ vdctrl0 = readl(host->base + VDCTRL0);
+ vdctrl2 = readl(host->base + VDCTRL2);
+ vdctrl3 = readl(host->base + VDCTRL3);
+ vdctrl4 = readl(host->base + VDCTRL4);
+
+ transfer_count = readl(host->base + host->devdata->transfer_count);
+
+ vmode.xres = TRANSFER_COUNT_GET_HCOUNT(transfer_count);
+ vmode.yres = TRANSFER_COUNT_GET_VCOUNT(transfer_count);
+
+ switch (CTRL_GET_WORD_LENGTH(ctrl)) {
+ case 0:
+ bits_per_pixel = 16;
+ break;
+ case 3:
+ bits_per_pixel = 32;
+ case 1:
+ default:
+ return -EINVAL;
+ }
+
+ fb_info->var.bits_per_pixel = bits_per_pixel;
+
+ vmode.pixclock = KHZ2PICOS(clk_get_rate(host->clk) / 1000U);
+ vmode.hsync_len = get_hsync_pulse_width(host, vdctrl2);
+ vmode.left_margin = GET_HOR_WAIT_CNT(vdctrl3) - vmode.hsync_len;
+ vmode.right_margin = VDCTRL2_GET_HSYNC_PERIOD(vdctrl2) - vmode.hsync_len -
+ vmode.left_margin - vmode.xres;
+ vmode.vsync_len = VDCTRL0_GET_VSYNC_PULSE_WIDTH(vdctrl0);
+ period = readl(host->base + VDCTRL1);
+ vmode.upper_margin = GET_VERT_WAIT_CNT(vdctrl3) - vmode.vsync_len;
+ vmode.lower_margin = period - vmode.vsync_len - vmode.upper_margin - vmode.yres;
+
+ vmode.vmode = FB_VMODE_NONINTERLACED;
+
+ vmode.sync = 0;
+ if (vdctrl0 & VDCTRL0_HSYNC_ACT_HIGH)
+ vmode.sync |= FB_SYNC_HOR_HIGH_ACT;
+ if (vdctrl0 & VDCTRL0_VSYNC_ACT_HIGH)
+ vmode.sync |= FB_SYNC_VERT_HIGH_ACT;
+
+ pr_debug("Reconstructed video mode:\n");
+ pr_debug("%dx%d, hsync: %u left: %u, right: %u, vsync: %u, upper: %u, lower: %u\n",
+ vmode.xres, vmode.yres,
+ vmode.hsync_len, vmode.left_margin, vmode.right_margin,
+ vmode.vsync_len, vmode.upper_margin, vmode.lower_margin);
+ pr_debug("pixclk: %ldkHz\n", PICOS2KHZ(vmode.pixclock));
+
+ fb_add_videomode(&vmode, &fb_info->modelist);
+
+ host->ld_intf_width = CTRL_GET_BUS_WIDTH(ctrl);
+ host->dotclk_delay = VDCTRL4_GET_DOTCLK_DLY(vdctrl4);
+
+ fb_info->fix.line_length = vmode.xres * (bits_per_pixel >> 3);
+
+ pa = readl(host->base + host->devdata->cur_buf);
+ fbsize = fb_info->fix.line_length * vmode.yres;
+ if (pa < fb_info->fix.smem_start)
+ return -EINVAL;
+ if (pa + fbsize > fb_info->fix.smem_start + fb_info->fix.smem_len)
+ return -EINVAL;
+ ofs = pa - fb_info->fix.smem_start;
+ if (ofs) {
+ memmove(fb_info->screen_base, fb_info->screen_base + ofs, fbsize);
+ writel(fb_info->fix.smem_start, host->base + host->devdata->next_buf);
+ }
+
+ line_count = fb_info->fix.smem_len / fb_info->fix.line_length;
+ fb_info->fix.ypanstep = 1;
+
+ clk_enable(host->clk);
+ host->enabled = 1;
+
+ return 0;
+}
+
+static int __devinit mxsfb_init_fbinfo(struct mxsfb_info *host)
+{
+ struct fb_info *fb_info = &host->fb_info;
+ struct fb_var_screeninfo *var = &fb_info->var;
+ struct mxsfb_platform_data *pdata = host->pdev->dev.platform_data;
+ dma_addr_t fb_phys;
+ void *fb_virt;
+ unsigned fb_size = pdata->fb_size;
+
+ fb_info->fbops = &mxsfb_ops;
+ fb_info->flags = FBINFO_FLAG_DEFAULT | FBINFO_READS_FAST;
+ strlcpy(fb_info->fix.id, "mxs", sizeof(fb_info->fix.id));
+ fb_info->fix.type = FB_TYPE_PACKED_PIXELS;
+ fb_info->fix.ypanstep = 1;
+ fb_info->fix.visual = FB_VISUAL_TRUECOLOR,
+ fb_info->fix.accel = FB_ACCEL_NONE;
+
+ var->bits_per_pixel = pdata->default_bpp ? pdata->default_bpp : 16;
+ var->nonstd = 0;
+ var->activate = FB_ACTIVATE_NOW;
+ var->accel_flags = 0;
+ var->vmode = FB_VMODE_NONINTERLACED;
+
+ host->dotclk_delay = pdata->dotclk_delay;
+ host->ld_intf_width = pdata->ld_intf_width;
+
+ /* Memory allocation for framebuffer */
+ if (pdata->fb_phys) {
+ if (!fb_size)
+ return -EINVAL;
+
+ fb_phys = pdata->fb_phys;
+
+ if (!request_mem_region(fb_phys, fb_size, host->pdev->name))
+ return -ENOMEM;
+
+ fb_virt = ioremap(fb_phys, fb_size);
+ if (!fb_virt) {
+ release_mem_region(fb_phys, fb_size);
+ return -ENOMEM;
+ }
+ host->mapped = 1;
+ } else {
+ if (!fb_size)
+ fb_size = SZ_2M; /* default */
+ fb_virt = alloc_pages_exact(fb_size, GFP_DMA);
+ if (!fb_info->screen_base)
+ return -ENOMEM;
+
+ fb_phys = virt_to_phys(fb_virt);
+ }
+
+ fb_info->fix.smem_start = fb_phys;
+ fb_info->screen_base = fb_virt;
+ fb_info->screen_size = fb_info->fix.smem_len = fb_size;
+
+ if (mxsfb_restore_mode(host))
+ memset(fb_virt, 0, fb_size);
+
+ return 0;
+}
+
+static void __devexit mxsfb_free_videomem(struct mxsfb_info *host)
+{
+ struct fb_info *fb_info = &host->fb_info;
+
+ if (host->mapped) {
+ iounmap(fb_info->screen_base);
+ release_mem_region(fb_info->fix.smem_start,
+ fb_info->screen_size);
+ } else {
+ free_pages_exact(fb_info->screen_base, fb_info->fix.smem_len);
+ }
+}
+
+static int __devinit mxsfb_probe(struct platform_device *pdev)
+{
+ struct mxsfb_platform_data *pdata = pdev->dev.platform_data;
+ struct resource *res;
+ struct mxsfb_info *host;
+ struct fb_info *fb_info;
+ struct fb_modelist *modelist;
+ int i, ret;
+
+ if (!pdata) {
+ dev_err(&pdev->dev, "No platformdata. Giving up\n");
+ return -ENODEV;
+ }
+
+ res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+ if (!res) {
+ dev_err(&pdev->dev, "Cannot get memory IO resource\n");
+ return -ENODEV;
+ }
+
+ if (!request_mem_region(res->start, resource_size(res), pdev->name))
+ return -EBUSY;
+
+ fb_info = framebuffer_alloc(sizeof(struct mxsfb_info), &pdev->dev);
+ if (!fb_info) {
+ dev_err(&pdev->dev, "Failed to allocate fbdev\n");
+ ret = -ENOMEM;
+ goto error_alloc_info;
+ }
+
+ host = to_imxfb_host(fb_info);
+
+ host->base = ioremap(res->start, resource_size(res));
+ if (!host->base) {
+ dev_err(&pdev->dev, "ioremap failed\n");
+ ret = -ENOMEM;
+ goto error_ioremap;
+ }
+
+ host->pdev = pdev;
+ platform_set_drvdata(pdev, host);
+
+ host->devdata = &mxsfb_devdata[pdev->id_entry->driver_data];
+
+ host->clk = clk_get(&host->pdev->dev, NULL);
+ if (IS_ERR(host->clk)) {
+ ret = PTR_ERR(host->clk);
+ goto error_getclock;
+ }
+
+ fb_info->pseudo_palette = kmalloc(sizeof(u32) * 16, GFP_KERNEL);
+ if (!fb_info->pseudo_palette) {
+ ret = -ENOMEM;
+ goto error_pseudo_pallette;
+ }
+
+ INIT_LIST_HEAD(&fb_info->modelist);
+
+ ret = mxsfb_init_fbinfo(host);
+ if (ret != 0)
+ goto error_init_fb;
+
+ for (i = 0; i < pdata->mode_count; i++)
+ fb_add_videomode(&pdata->mode_list[i], &fb_info->modelist);
+
+ modelist = list_first_entry(&fb_info->modelist,
+ struct fb_modelist, list);
+ fb_videomode_to_var(&fb_info->var, &modelist->mode);
+
+ /* init the color fields */
+ mxsfb_check_var(&fb_info->var, fb_info);
+
+ platform_set_drvdata(pdev, fb_info);
+
+ ret = register_framebuffer(fb_info);
+ if (ret != 0) {
+ dev_err(&pdev->dev,"Failed to register framebuffer\n");
+ goto error_register;
+ }
+
+ if (!host->enabled) {
+ writel(0, host->base + CTRL);
+ mxsfb_set_par(fb_info);
+ mxsfb_enable_controller(fb_info);
+ }
+
+ return 0;
+
+error_register:
+ if (host->enabled)
+ clk_disable(host->clk);
+ fb_destroy_modelist(&fb_info->modelist);
+error_init_fb:
+ kfree(fb_info->pseudo_palette);
+error_pseudo_pallette:
+ clk_put(host->clk);
+error_getclock:
+ iounmap(host->base);
+error_ioremap:
+ framebuffer_release(fb_info);
+error_alloc_info:
+ release_mem_region(res->start, resource_size(res));
+
+ return ret;
+}
+
+static int __devexit mxsfb_remove(struct platform_device *pdev)
+{
+ struct fb_info *fb_info = platform_get_drvdata(pdev);
+ struct mxsfb_info *host = to_imxfb_host(fb_info);
+ struct resource *res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+
+ mxsfb_disable_controller(fb_info);
+
+ unregister_framebuffer(fb_info);
+ kfree(fb_info->pseudo_palette);
+ mxsfb_free_videomem(host);
+ iounmap(host->base);
+ clk_put(host->clk);
+
+ framebuffer_release(fb_info);
+ release_mem_region(res->start, resource_size(res));
+
+ platform_set_drvdata(pdev, NULL);
+
+ return 0;
+}
+
+static struct platform_device_id mxsfb_devtype[] = {
+ {
+ .name = "imx23-fb",
+ .driver_data = MXSFB_MX23,
+ }, {
+ .name = "imx28-fb",
+ .driver_data = MXSFB_MX28,
+ }, {
+ },
+};
+MODULE_DEVICE_TABLE(platform, mxsfb_devtype);
+
+static struct platform_driver mxsfb_driver = {
+ .probe = mxsfb_probe,
+ .remove = __devexit_p(mxsfb_remove),
+ .id_table = mxsfb_devtype,
+ .driver = {
+ .name = DRIVER_NAME,
+ },
+};
+
+static int __init mxsfb_init(void)
+{
+ return platform_driver_register(&mxsfb_driver);
+}
+
+static void __exit mxsfb_exit(void)
+{
+ platform_driver_unregister(&mxsfb_driver);
+}
+
+module_init(mxsfb_init);
+module_exit(mxsfb_exit);
+
+MODULE_DESCRIPTION("Freescale mxs framebuffer driver");
+MODULE_AUTHOR("Sascha Hauer, Pengutronix");
+MODULE_LICENSE("GPL");
--
1.7.2.3
^ permalink raw reply related [flat|nested] 65+ messages in thread
* [PATCH 1/2] video: Add i.MX23/28 framebuffer driver
@ 2011-02-09 13:20 ` Sascha Hauer
0 siblings, 0 replies; 65+ messages in thread
From: Sascha Hauer @ 2011-02-09 13:20 UTC (permalink / raw)
To: linux-arm-kernel
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Cc: Paul Mundt <lethal@linux-sh.org>
Cc: linux-fbdev at vger.kernel.org
Cc: Shawn Guo <shawn.guo@freescale.com>
---
arch/arm/mach-mxs/include/mach/fb.h | 48 ++
drivers/video/Kconfig | 9 +
drivers/video/Makefile | 1 +
drivers/video/mxsfb.c | 914 +++++++++++++++++++++++++++++++++++
4 files changed, 972 insertions(+), 0 deletions(-)
create mode 100644 arch/arm/mach-mxs/include/mach/fb.h
create mode 100644 drivers/video/mxsfb.c
diff --git a/arch/arm/mach-mxs/include/mach/fb.h b/arch/arm/mach-mxs/include/mach/fb.h
new file mode 100644
index 0000000..923f397
--- /dev/null
+++ b/arch/arm/mach-mxs/include/mach/fb.h
@@ -0,0 +1,48 @@
+/*
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation; either version 2
+ * of the License, or (at your option) any later version.
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
+ * MA 02110-1301, USA.
+ */
+
+#ifndef __MACH_FB_H
+#define __MACH_FB_H
+
+#include <linux/fb.h>
+
+#define STMLCDIF_8BIT 1 /** pixel data bus to the display is of 8 bit width */
+#define STMLCDIF_16BIT 0 /** pixel data bus to the display is of 16 bit width */
+#define STMLCDIF_18BIT 2 /** pixel data bus to the display is of 18 bit width */
+#define STMLCDIF_24BIT 3 /** pixel data bus to the display is of 24 bit width */
+
+#define FB_SYNC_DATA_ENABLE_HIGH_ACT 64
+
+struct mxsfb_platform_data {
+ struct fb_videomode *mode_list;
+ unsigned mode_count;
+
+ unsigned default_bpp;
+
+ unsigned dotclk_delay; /* refer manual HW_LCDIF_VDCTRL4 register */
+ unsigned ld_intf_width; /* refer STMLCDIF_* macros */
+
+ unsigned fb_size; /* Size of the video memory. If zero a
+ * default will be used
+ */
+ unsigned long fb_phys; /* physical address for the video memory. If
+ * zero the framebuffer memory will be dynamically
+ * allocated. If specified,fb_size must also be specified.
+ * fb_phys must be unused by Linux.
+ */
+};
+
+#endif /* __MACH_FB_H */
diff --git a/drivers/video/Kconfig b/drivers/video/Kconfig
index 6bafb51b..e0ea23f 100644
--- a/drivers/video/Kconfig
+++ b/drivers/video/Kconfig
@@ -2365,6 +2365,15 @@ config FB_JZ4740
help
Framebuffer support for the JZ4740 SoC.
+config FB_MXS
+ tristate "MXS LCD framebuffer support"
+ depends on FB && ARCH_MXS
+ select FB_CFB_FILLRECT
+ select FB_CFB_COPYAREA
+ select FB_CFB_IMAGEBLIT
+ help
+ Framebuffer support for the MXS SoC.
+
source "drivers/video/omap/Kconfig"
source "drivers/video/omap2/Kconfig"
diff --git a/drivers/video/Makefile b/drivers/video/Makefile
index 8c8fabd..9a096ae 100644
--- a/drivers/video/Makefile
+++ b/drivers/video/Makefile
@@ -153,6 +153,7 @@ obj-$(CONFIG_FB_BFIN_T350MCQB) += bfin-t350mcqb-fb.o
obj-$(CONFIG_FB_BFIN_7393) += bfin_adv7393fb.o
obj-$(CONFIG_FB_MX3) += mx3fb.o
obj-$(CONFIG_FB_DA8XX) += da8xx-fb.o
+obj-$(CONFIG_FB_MXS) += mxsfb.o
# the test framebuffer is last
obj-$(CONFIG_FB_VIRTUAL) += vfb.o
diff --git a/drivers/video/mxsfb.c b/drivers/video/mxsfb.c
new file mode 100644
index 0000000..92a5804
--- /dev/null
+++ b/drivers/video/mxsfb.c
@@ -0,0 +1,914 @@
+/*
+ * Copyright (C) 2010 Juergen Beisert, Pengutronix
+ *
+ * This code is based on:
+ * Author: Vitaly Wool <vital@embeddedalley.com>
+ *
+ * Copyright 2008-2009 Freescale Semiconductor, Inc. All Rights Reserved.
+ * Copyright 2008 Embedded Alley Solutions, Inc All Rights Reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation; either version 2
+ * of the License, or (at your option) any later version.
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ */
+
+#define DRIVER_NAME "mxsfb"
+
+/**
+ * @file
+ * @brief LCDIF driver for i.MX23 and i.MX28
+ *
+ * The LCDIF support four modes of operation
+ * - MPU interface (to drive smart displays) -> not supported yet
+ * - VSYNC interface (like MPU interface plus Vsync) -> not supported yet
+ * - Dotclock interface (to drive LC displays with RGB data and sync signals)
+ * - DVI (to drive ITU-R BT656) -> not supported yet
+ *
+ * This driver depends on a correct setup of the pins used for this purpose
+ * (platform specific).
+ *
+ * For the developer: Don't forget to set the data bus width to the display
+ * in the imx_fb_videomode structure. You will else end up with ugly colours.
+ * If you fight against jitter you can vary the clock delay. This is a feature
+ * of the i.MX28 and you can vary it between 2 ns ... 8 ns in 2 ns steps. Give
+ * the required value in the imx_fb_videomode structure.
+ */
+
+#include <linux/kernel.h>
+#include <linux/platform_device.h>
+#include <linux/clk.h>
+#include <linux/dma-mapping.h>
+#include <linux/io.h>
+#include <mach/fb.h>
+#include <mach/hardware.h>
+#include <mach/mxs.h>
+
+#define CTRL 0x00
+#define CTRL1 0x10
+#define MX28_CTRL2 0x20
+#define MX23_TRANSFER_COUNT 0x20
+#define MX28_TRANSFER_COUNT 0x30
+#define MX28_CUR_BUF 0x40
+#define MX28_NEXT_BUF 0x50
+#define MX23_CUR_BUF 0x30
+#define MX23_NEXT_BUF 0x40
+#define TIMING 0x60
+#define VDCTRL0 0x70
+#define VDCTRL1 0x80
+#define VDCTRL2 0x90
+#define VDCTRL3 0xa0
+#define VDCTRL4 0xb0
+#define DVICTRL0 0xc0
+#define DVICTRL1 0xd0
+#define DVICTRL2 0xe0
+#define DVICTRL3 0xf0
+#define DVICTRL4 0x100
+#define MX28_DATA 0x180
+#define MX23_DATA 0x1b0
+#define MX28_DEBUG0 0x1d0
+#define MX23_DEBUG0 0x1f0
+
+#define CTRL_SFTRST (1 << 31)
+#define CTRL_CLKGATE (1 << 30)
+#define CTRL_BYPASS_COUNT (1 << 19)
+#define CTRL_VSYNC_MODE (1 << 18)
+#define CTRL_DOTCLK_MODE (1 << 17)
+#define CTRL_DATA_SELECT (1 << 16)
+#define CTRL_SET_BUS_WIDTH(x) (((x) & 0x3) << 10)
+#define CTRL_GET_BUS_WIDTH(x) (((x) >> 10) & 0x3)
+#define CTRL_SET_WORD_LENGTH(x) (((x) & 0x3) << 8)
+#define CTRL_GET_WORD_LENGTH(x) (((x) >> 8) & 0x3)
+#define CTRL_MASTER (1 << 5)
+#define CTRL_DF16 (1 << 3)
+#define CTRL_DF18 (1 << 2)
+#define CTRL_DF24 (1 << 1)
+#define CTRL_RUN (1 << 0)
+
+#define CTRL1_FIFO_CLEAR (1 << 21)
+#define CTRL1_SET_BYTE_PACKAGING(x) (((x) & 0xf) << 16)
+#define CTRL1_GET_BYTE_PACKAGING(x) (((x) >> 16) & 0xf)
+
+#define TRANSFER_COUNT_SET_VCOUNT(x) (((x) & 0xffff) << 16)
+#define TRANSFER_COUNT_GET_VCOUNT(x) (((x) >> 16) & 0xffff)
+#define TRANSFER_COUNT_SET_HCOUNT(x) ((x) & 0xffff)
+#define TRANSFER_COUNT_GET_HCOUNT(x) ((x) & 0xffff)
+
+
+#define VDCTRL0_ENABLE_PRESENT (1 << 28)
+#define VDCTRL0_VSYNC_ACT_HIGH (1 << 27)
+#define VDCTRL0_HSYNC_ACT_HIGH (1 << 26)
+#define VDCTRL0_DOTCLK_ACTIVE_HIGH (1 << 25)
+#define VDCTRL0_POL_ACTIVE_HIGH (1 << 24)
+#define VDCTRL0_VSYNC_PERIOD_UNIT (1 << 21)
+#define VDCTRL0_VSYNC_PULSE_WIDTH_UNIT (1 << 20)
+#define VDCTRL0_HALF_LINE (1 << 19)
+#define VDCTRL0_HALF_LINE_MODE (1 << 18)
+#define VDCTRL0_SET_VSYNC_PULSE_WIDTH(x) ((x) & 0x3ffff)
+#define VDCTRL0_GET_VSYNC_PULSE_WIDTH(x) ((x) & 0x3ffff)
+
+#define VDCTRL2_SET_HSYNC_PERIOD(x) ((x) & 0x3ffff)
+#define VDCTRL2_GET_HSYNC_PERIOD(x) ((x) & 0x3ffff)
+
+#define VDCTRL3_MUX_SYNC_SIGNALS (1 << 29)
+#define VDCTRL3_VSYNC_ONLY (1 << 28)
+#define SET_HOR_WAIT_CNT(x) (((x) & 0xfff) << 16)
+#define GET_HOR_WAIT_CNT(x) (((x) >> 16) & 0xfff)
+#define SET_VERT_WAIT_CNT(x) ((x) & 0xffff)
+#define GET_VERT_WAIT_CNT(x) ((x) & 0xffff)
+
+#define VDCTRL4_SET_DOTCLK_DLY(x) (((x) & 0x7) << 29) /* i.MX28 only */
+#define VDCTRL4_GET_DOTCLK_DLY(x) (((x) >> 29) & 0x7) /* i.MX28 only */
+#define VDCTRL4_SYNC_SIGNALS_ON (1 << 18)
+#define SET_DOTCLK_H_VALID_DATA_CNT(x) ((x) & 0x3ffff)
+
+#define DEBUG0_HSYNC (1 < 26)
+#define DEBUG0_VSYNC (1 < 25)
+
+#define MIN_XRES 120
+#define MIN_YRES 120
+
+#define RED 0
+#define GREEN 1
+#define BLUE 2
+#define TRANSP 3
+
+enum mxsfb_devtype {
+ MXSFB_MX23,
+ MXSFB_MX28,
+};
+
+/* CPU dependent register offsets */
+struct mxsfb_devdata {
+ unsigned transfer_count;
+ unsigned cur_buf;
+ unsigned next_buf;
+ unsigned debug0;
+ unsigned hs_wdth_mask;
+ unsigned hs_wdth_shift;
+};
+
+struct mxsfb_info {
+ struct fb_info fb_info;
+ struct platform_device *pdev;
+ struct clk *clk;
+ void __iomem *base; /* registers */
+ unsigned allocated_size;
+ int enabled;
+ unsigned ld_intf_width;
+ unsigned dotclk_delay;
+ const struct mxsfb_devdata *devdata;
+ int mapped;
+};
+
+static const struct mxsfb_devdata mxsfb_devdata[] = {
+ [MXSFB_MX23] = {
+ .transfer_count = MX23_TRANSFER_COUNT,
+ .cur_buf = MX23_CUR_BUF,
+ .next_buf = MX23_NEXT_BUF,
+ .debug0 = MX23_DEBUG0,
+ .hs_wdth_mask = 0xff,
+ .hs_wdth_shift = 24,
+ },
+ [MXSFB_MX28] = {
+ .transfer_count = MX28_TRANSFER_COUNT,
+ .cur_buf = MX28_CUR_BUF,
+ .next_buf = MX28_NEXT_BUF,
+ .debug0 = MX28_DEBUG0,
+ .hs_wdth_mask = 0x3fff,
+ .hs_wdth_shift = 18,
+ },
+};
+
+#define to_imxfb_host(x) (container_of(x, struct mxsfb_info, fb_info))
+
+/* mask and shift depends on architecture */
+static inline u32 set_hsync_pulse_width(struct mxsfb_info *host, unsigned val)
+{
+ return (val & host->devdata->hs_wdth_mask) <<
+ host->devdata->hs_wdth_shift;
+}
+
+static inline u32 get_hsync_pulse_width(struct mxsfb_info *host, unsigned val)
+{
+ return (val >> host->devdata->hs_wdth_shift) &
+ host->devdata->hs_wdth_mask;
+}
+
+static const struct fb_bitfield def_rgb565[] = {
+ [RED] = {
+ .offset = 11,
+ .length = 5,
+ },
+ [GREEN] = {
+ .offset = 5,
+ .length = 6,
+ },
+ [BLUE] = {
+ .offset = 0,
+ .length = 5,
+ },
+ [TRANSP] = { /* no support for transparency */
+ .length = 0,
+ }
+};
+
+static const struct fb_bitfield def_rgb666[] = {
+ [RED] = {
+ .offset = 16,
+ .length = 6,
+ },
+ [GREEN] = {
+ .offset = 8,
+ .length = 6,
+ },
+ [BLUE] = {
+ .offset = 0,
+ .length = 6,
+ },
+ [TRANSP] = { /* no support for transparency */
+ .length = 0,
+ }
+};
+
+static const struct fb_bitfield def_rgb888[] = {
+ [RED] = {
+ .offset = 16,
+ .length = 8,
+ },
+ [GREEN] = {
+ .offset = 8,
+ .length = 8,
+ },
+ [BLUE] = {
+ .offset = 0,
+ .length = 8,
+ },
+ [TRANSP] = { /* no support for transparency */
+ .length = 0,
+ }
+};
+
+static inline unsigned chan_to_field(unsigned chan, struct fb_bitfield *bf)
+{
+ chan &= 0xffff;
+ chan >>= 16 - bf->length;
+ return chan << bf->offset;
+}
+
+static int mxsfb_check_var(struct fb_var_screeninfo *var,
+ struct fb_info *fb_info)
+{
+ struct mxsfb_info *host = to_imxfb_host(fb_info);
+ const struct fb_bitfield *rgb = NULL;
+
+ if (var->xres < MIN_XRES)
+ var->xres = MIN_XRES;
+ if (var->yres < MIN_YRES)
+ var->yres = MIN_YRES;
+
+ var->xres_virtual = var->xres;
+
+ var->yres_virtual = var->yres;
+
+ switch (var->bits_per_pixel) {
+ case 16:
+ /* always expect RGB 565 */
+ rgb = def_rgb565;
+ break;
+ case 32:
+ switch (host->ld_intf_width) {
+ case STMLCDIF_8BIT:
+ pr_debug("Unsupported LCD bus width mapping\n");
+ break;
+ case STMLCDIF_16BIT:
+ case STMLCDIF_18BIT:
+ /* 24 bit to 18 bit mapping */
+ rgb = def_rgb666;
+ break;
+ case STMLCDIF_24BIT:
+ /* real 24 bit */
+ rgb = def_rgb888;
+ break;
+ }
+ break;
+ default:
+ pr_debug("Unsupported colour depth: %u\n", var->bits_per_pixel);
+ return -EINVAL;
+ }
+
+ /*
+ * Copy the RGB parameters for this display
+ * from the machine specific parameters.
+ */
+ var->red = rgb[RED];
+ var->green = rgb[GREEN];
+ var->blue = rgb[BLUE];
+ var->transp = rgb[TRANSP];
+
+ return 0;
+}
+
+static void mxsfb_enable_controller(struct fb_info *fb_info)
+{
+ struct mxsfb_info *host = to_imxfb_host(fb_info);
+ u32 reg;
+
+ dev_dbg(&host->pdev->dev, "%s\n", __func__);
+
+ clk_enable(host->clk);
+ clk_set_rate(host->clk, PICOS2KHZ(fb_info->var.pixclock) * 1000U);
+
+ /* if it was disabled, re-enable the mode again */
+ reg = readl(host->base + CTRL);
+ reg |= CTRL_DOTCLK_MODE;
+ writel(reg, host->base + CTRL);
+
+ /* enable the SYNC signals first, then the DMA engine */
+ reg = readl(host->base + VDCTRL4);
+ reg |= VDCTRL4_SYNC_SIGNALS_ON;
+ writel(reg, host->base + VDCTRL4);
+
+ reg = readl(host->base + CTRL);
+ reg |= CTRL_RUN;
+ writel(reg, host->base + CTRL);
+
+ host->enabled = 1;
+}
+
+static void mxsfb_disable_controller(struct fb_info *fb_info)
+{
+ struct mxsfb_info *host = to_imxfb_host(fb_info);
+ unsigned loop;
+ u32 reg;
+
+ dev_dbg(&host->pdev->dev, "%s\n", __func__);
+
+ /*
+ * Even if we disable the controller here, it will still continue
+ * until its FIFOs are running out of data
+ */
+ reg = readl(host->base + CTRL);
+ reg &= ~CTRL_DOTCLK_MODE;
+ writel(reg, host->base + CTRL);
+
+ loop = 1000;
+ while (loop) {
+ reg = readl(host->base + CTRL);
+ if (!(reg & CTRL_RUN))
+ break;
+ loop--;
+ }
+
+ reg = readl(host->base + VDCTRL4);
+ reg &= ~VDCTRL4_SYNC_SIGNALS_ON;
+ writel(reg, host->base + VDCTRL4);
+
+ clk_disable(host->clk);
+
+ host->enabled = 0;
+}
+
+static int mxsfb_set_par(struct fb_info *fb_info)
+{
+ struct mxsfb_info *host = to_imxfb_host(fb_info);
+ u32 ctrl, vdctrl0, vdctrl4;
+ int line_size, fb_size;
+ int reenable = 0;
+
+ line_size = fb_info->var.xres * (fb_info->var.bits_per_pixel >> 3);
+ fb_size = fb_info->var.yres_virtual * line_size;
+
+ if (fb_size > fb_info->fix.smem_len)
+ return -ENOMEM;
+
+ fb_info->fix.line_length = line_size;
+
+ /*
+ * It seems, you can't re-program the controller if it is still running.
+ * This may lead into shifted pictures (FIFO issue?).
+ * So, first stop the controller and drain its FIFOs
+ */
+ if (host->enabled) {
+ reenable = 1;
+ mxsfb_disable_controller(fb_info);
+ }
+
+ /* clear the FIFOs */
+ writel(CTRL1_FIFO_CLEAR, host->base + CTRL1 + 4);
+
+ ctrl = CTRL_BYPASS_COUNT | CTRL_MASTER |
+ CTRL_SET_BUS_WIDTH(host->ld_intf_width);;
+
+ switch (fb_info->var.bits_per_pixel) {
+ case 16:
+ dev_dbg(&host->pdev->dev, "Setting up RGB565 mode\n");
+ ctrl |= CTRL_SET_WORD_LENGTH(0);
+ writel(CTRL1_SET_BYTE_PACKAGING(0xf), host->base + CTRL1);
+ break;
+ case 32:
+ dev_dbg(&host->pdev->dev, "Setting up RGB888/666 mode\n");
+ ctrl |= CTRL_SET_WORD_LENGTH(3);
+ switch (host->ld_intf_width) {
+ case STMLCDIF_8BIT:
+ dev_dbg(&host->pdev->dev,
+ "Unsupported LCD bus width mapping\n");
+ return -EINVAL;
+ case STMLCDIF_16BIT:
+ case STMLCDIF_18BIT:
+ /* 24 bit to 18 bit mapping */
+ ctrl |= CTRL_DF24; /* ignore the upper 2 bits in
+ * each colour component
+ */
+ break;
+ case STMLCDIF_24BIT:
+ /* real 24 bit */
+ break;
+ }
+ /* do not use packed pixels = one pixel per word instead */
+ writel(CTRL1_SET_BYTE_PACKAGING(0x7), host->base + CTRL1);
+ break;
+ default:
+ dev_dbg(&host->pdev->dev, "Unhandled color depth of %u\n",
+ fb_info->var.bits_per_pixel);
+ return -EINVAL;
+ }
+
+ writel(ctrl, host->base + CTRL);
+
+ writel(TRANSFER_COUNT_SET_VCOUNT(fb_info->var.yres) |
+ TRANSFER_COUNT_SET_HCOUNT(fb_info->var.xres),
+ host->base + host->devdata->transfer_count);
+
+ vdctrl0 = VDCTRL0_ENABLE_PRESENT | /* always in DOTCLOCK mode */
+ VDCTRL0_VSYNC_PERIOD_UNIT |
+ VDCTRL0_VSYNC_PULSE_WIDTH_UNIT |
+ VDCTRL0_SET_VSYNC_PULSE_WIDTH(fb_info->var.vsync_len);
+ if (fb_info->var.sync & FB_SYNC_HOR_HIGH_ACT)
+ vdctrl0 |= VDCTRL0_HSYNC_ACT_HIGH;
+ if (fb_info->var.sync & FB_SYNC_VERT_HIGH_ACT)
+ vdctrl0 |= VDCTRL0_VSYNC_ACT_HIGH;
+ if (fb_info->var.sync & FB_SYNC_DATA_ENABLE_HIGH_ACT)
+ vdctrl0 |= VDCTRL0_POL_ACTIVE_HIGH;
+
+ writel(vdctrl0, host->base + VDCTRL0);
+
+ /* frame length in lines */
+ writel(fb_info->var.upper_margin + fb_info->var.vsync_len +
+ fb_info->var.lower_margin + fb_info->var.yres,
+ host->base + VDCTRL1);
+
+ /* line length in units of clocks or pixels */
+ writel(set_hsync_pulse_width(host, fb_info->var.hsync_len) |
+ VDCTRL2_SET_HSYNC_PERIOD(fb_info->var.left_margin +
+ fb_info->var.hsync_len + fb_info->var.right_margin +
+ fb_info->var.xres),
+ host->base + VDCTRL2);
+
+ writel(SET_HOR_WAIT_CNT(fb_info->var.left_margin +
+ fb_info->var.hsync_len) |
+ SET_VERT_WAIT_CNT(fb_info->var.upper_margin +
+ fb_info->var.vsync_len),
+ host->base + VDCTRL3);
+
+ vdctrl4 = SET_DOTCLK_H_VALID_DATA_CNT(fb_info->var.xres);
+ if (cpu_is_mx28())
+ vdctrl4 |= VDCTRL4_SET_DOTCLK_DLY(host->dotclk_delay);
+ writel(vdctrl4, host->base + VDCTRL4);
+
+ writel(fb_info->fix.smem_start +
+ fb_info->fix.line_length * fb_info->var.yoffset,
+ host->base + host->devdata->next_buf);
+
+ if (reenable)
+ mxsfb_enable_controller(fb_info);
+
+ return 0;
+}
+
+static int mxsfb_setcolreg(u_int regno, u_int red, u_int green, u_int blue,
+ u_int transp, struct fb_info *fb_info)
+{
+ unsigned int val;
+ int ret = -EINVAL;
+
+ /*
+ * If greyscale is true, then we convert the RGB value
+ * to greyscale no matter what visual we are using.
+ */
+ if (fb_info->var.grayscale)
+ red = green = blue = (19595 * red + 38470 * green +
+ 7471 * blue) >> 16;
+
+ switch (fb_info->fix.visual) {
+ case FB_VISUAL_TRUECOLOR:
+ /*
+ * 12 or 16-bit True Colour. We encode the RGB value
+ * according to the RGB bitfield information.
+ */
+ if (regno < 16) {
+ u32 *pal = fb_info->pseudo_palette;
+
+ val = chan_to_field(red, &fb_info->var.red);
+ val |= chan_to_field(green, &fb_info->var.green);
+ val |= chan_to_field(blue, &fb_info->var.blue);
+
+ pal[regno] = val;
+ ret = 0;
+ }
+ break;
+
+ case FB_VISUAL_STATIC_PSEUDOCOLOR:
+ case FB_VISUAL_PSEUDOCOLOR:
+ break;
+ }
+
+ return ret;
+}
+
+static int mxsfb_blank(int blank, struct fb_info *fb_info)
+{
+ struct mxsfb_info *host = to_imxfb_host(fb_info);
+
+ switch (blank) {
+ case FB_BLANK_POWERDOWN:
+ case FB_BLANK_VSYNC_SUSPEND:
+ case FB_BLANK_HSYNC_SUSPEND:
+ case FB_BLANK_NORMAL:
+ if (host->enabled)
+ mxsfb_disable_controller(fb_info);
+ break;
+
+ case FB_BLANK_UNBLANK:
+ if (!host->enabled)
+ mxsfb_enable_controller(fb_info);
+ break;
+ }
+ return 0;
+}
+
+static int mxsfb_pan_display(struct fb_var_screeninfo *var,
+ struct fb_info *fb_info)
+{
+ struct mxsfb_info *host = to_imxfb_host(fb_info);
+ unsigned offset;
+
+ if (var->xoffset != 0)
+ return -EINVAL;
+
+ offset = fb_info->fix.line_length * var->yoffset;
+
+ /* update on next VSYNC */
+ writel(fb_info->fix.smem_start + offset,
+ host->base + host->devdata->next_buf);
+
+ return 0;
+}
+
+static struct fb_ops mxsfb_ops = {
+ .owner = THIS_MODULE,
+ .fb_check_var = mxsfb_check_var,
+ .fb_set_par = mxsfb_set_par,
+ .fb_setcolreg = mxsfb_setcolreg,
+ .fb_blank = mxsfb_blank,
+ .fb_pan_display = mxsfb_pan_display,
+ .fb_fillrect = cfb_fillrect,
+ .fb_copyarea = cfb_copyarea,
+ .fb_imageblit = cfb_imageblit,
+};
+
+static int __devinit mxsfb_restore_mode(struct mxsfb_info *host)
+{
+ struct fb_info *fb_info = &host->fb_info;
+ unsigned line_count;
+ unsigned period;
+ unsigned long pa, fbsize;
+ int bits_per_pixel, ofs;
+ u32 transfer_count, vdctrl0, vdctrl2, vdctrl3, vdctrl4, ctrl;
+ struct fb_videomode vmode;
+
+ /* Only restore the mode when the controller is running */
+ ctrl = readl(host->base + CTRL);
+ if (!(ctrl & CTRL_RUN))
+ return -EINVAL;
+
+ vdctrl0 = readl(host->base + VDCTRL0);
+ vdctrl2 = readl(host->base + VDCTRL2);
+ vdctrl3 = readl(host->base + VDCTRL3);
+ vdctrl4 = readl(host->base + VDCTRL4);
+
+ transfer_count = readl(host->base + host->devdata->transfer_count);
+
+ vmode.xres = TRANSFER_COUNT_GET_HCOUNT(transfer_count);
+ vmode.yres = TRANSFER_COUNT_GET_VCOUNT(transfer_count);
+
+ switch (CTRL_GET_WORD_LENGTH(ctrl)) {
+ case 0:
+ bits_per_pixel = 16;
+ break;
+ case 3:
+ bits_per_pixel = 32;
+ case 1:
+ default:
+ return -EINVAL;
+ }
+
+ fb_info->var.bits_per_pixel = bits_per_pixel;
+
+ vmode.pixclock = KHZ2PICOS(clk_get_rate(host->clk) / 1000U);
+ vmode.hsync_len = get_hsync_pulse_width(host, vdctrl2);
+ vmode.left_margin = GET_HOR_WAIT_CNT(vdctrl3) - vmode.hsync_len;
+ vmode.right_margin = VDCTRL2_GET_HSYNC_PERIOD(vdctrl2) - vmode.hsync_len -
+ vmode.left_margin - vmode.xres;
+ vmode.vsync_len = VDCTRL0_GET_VSYNC_PULSE_WIDTH(vdctrl0);
+ period = readl(host->base + VDCTRL1);
+ vmode.upper_margin = GET_VERT_WAIT_CNT(vdctrl3) - vmode.vsync_len;
+ vmode.lower_margin = period - vmode.vsync_len - vmode.upper_margin - vmode.yres;
+
+ vmode.vmode = FB_VMODE_NONINTERLACED;
+
+ vmode.sync = 0;
+ if (vdctrl0 & VDCTRL0_HSYNC_ACT_HIGH)
+ vmode.sync |= FB_SYNC_HOR_HIGH_ACT;
+ if (vdctrl0 & VDCTRL0_VSYNC_ACT_HIGH)
+ vmode.sync |= FB_SYNC_VERT_HIGH_ACT;
+
+ pr_debug("Reconstructed video mode:\n");
+ pr_debug("%dx%d, hsync: %u left: %u, right: %u, vsync: %u, upper: %u, lower: %u\n",
+ vmode.xres, vmode.yres,
+ vmode.hsync_len, vmode.left_margin, vmode.right_margin,
+ vmode.vsync_len, vmode.upper_margin, vmode.lower_margin);
+ pr_debug("pixclk: %ldkHz\n", PICOS2KHZ(vmode.pixclock));
+
+ fb_add_videomode(&vmode, &fb_info->modelist);
+
+ host->ld_intf_width = CTRL_GET_BUS_WIDTH(ctrl);
+ host->dotclk_delay = VDCTRL4_GET_DOTCLK_DLY(vdctrl4);
+
+ fb_info->fix.line_length = vmode.xres * (bits_per_pixel >> 3);
+
+ pa = readl(host->base + host->devdata->cur_buf);
+ fbsize = fb_info->fix.line_length * vmode.yres;
+ if (pa < fb_info->fix.smem_start)
+ return -EINVAL;
+ if (pa + fbsize > fb_info->fix.smem_start + fb_info->fix.smem_len)
+ return -EINVAL;
+ ofs = pa - fb_info->fix.smem_start;
+ if (ofs) {
+ memmove(fb_info->screen_base, fb_info->screen_base + ofs, fbsize);
+ writel(fb_info->fix.smem_start, host->base + host->devdata->next_buf);
+ }
+
+ line_count = fb_info->fix.smem_len / fb_info->fix.line_length;
+ fb_info->fix.ypanstep = 1;
+
+ clk_enable(host->clk);
+ host->enabled = 1;
+
+ return 0;
+}
+
+static int __devinit mxsfb_init_fbinfo(struct mxsfb_info *host)
+{
+ struct fb_info *fb_info = &host->fb_info;
+ struct fb_var_screeninfo *var = &fb_info->var;
+ struct mxsfb_platform_data *pdata = host->pdev->dev.platform_data;
+ dma_addr_t fb_phys;
+ void *fb_virt;
+ unsigned fb_size = pdata->fb_size;
+
+ fb_info->fbops = &mxsfb_ops;
+ fb_info->flags = FBINFO_FLAG_DEFAULT | FBINFO_READS_FAST;
+ strlcpy(fb_info->fix.id, "mxs", sizeof(fb_info->fix.id));
+ fb_info->fix.type = FB_TYPE_PACKED_PIXELS;
+ fb_info->fix.ypanstep = 1;
+ fb_info->fix.visual = FB_VISUAL_TRUECOLOR,
+ fb_info->fix.accel = FB_ACCEL_NONE;
+
+ var->bits_per_pixel = pdata->default_bpp ? pdata->default_bpp : 16;
+ var->nonstd = 0;
+ var->activate = FB_ACTIVATE_NOW;
+ var->accel_flags = 0;
+ var->vmode = FB_VMODE_NONINTERLACED;
+
+ host->dotclk_delay = pdata->dotclk_delay;
+ host->ld_intf_width = pdata->ld_intf_width;
+
+ /* Memory allocation for framebuffer */
+ if (pdata->fb_phys) {
+ if (!fb_size)
+ return -EINVAL;
+
+ fb_phys = pdata->fb_phys;
+
+ if (!request_mem_region(fb_phys, fb_size, host->pdev->name))
+ return -ENOMEM;
+
+ fb_virt = ioremap(fb_phys, fb_size);
+ if (!fb_virt) {
+ release_mem_region(fb_phys, fb_size);
+ return -ENOMEM;
+ }
+ host->mapped = 1;
+ } else {
+ if (!fb_size)
+ fb_size = SZ_2M; /* default */
+ fb_virt = alloc_pages_exact(fb_size, GFP_DMA);
+ if (!fb_info->screen_base)
+ return -ENOMEM;
+
+ fb_phys = virt_to_phys(fb_virt);
+ }
+
+ fb_info->fix.smem_start = fb_phys;
+ fb_info->screen_base = fb_virt;
+ fb_info->screen_size = fb_info->fix.smem_len = fb_size;
+
+ if (mxsfb_restore_mode(host))
+ memset(fb_virt, 0, fb_size);
+
+ return 0;
+}
+
+static void __devexit mxsfb_free_videomem(struct mxsfb_info *host)
+{
+ struct fb_info *fb_info = &host->fb_info;
+
+ if (host->mapped) {
+ iounmap(fb_info->screen_base);
+ release_mem_region(fb_info->fix.smem_start,
+ fb_info->screen_size);
+ } else {
+ free_pages_exact(fb_info->screen_base, fb_info->fix.smem_len);
+ }
+}
+
+static int __devinit mxsfb_probe(struct platform_device *pdev)
+{
+ struct mxsfb_platform_data *pdata = pdev->dev.platform_data;
+ struct resource *res;
+ struct mxsfb_info *host;
+ struct fb_info *fb_info;
+ struct fb_modelist *modelist;
+ int i, ret;
+
+ if (!pdata) {
+ dev_err(&pdev->dev, "No platformdata. Giving up\n");
+ return -ENODEV;
+ }
+
+ res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+ if (!res) {
+ dev_err(&pdev->dev, "Cannot get memory IO resource\n");
+ return -ENODEV;
+ }
+
+ if (!request_mem_region(res->start, resource_size(res), pdev->name))
+ return -EBUSY;
+
+ fb_info = framebuffer_alloc(sizeof(struct mxsfb_info), &pdev->dev);
+ if (!fb_info) {
+ dev_err(&pdev->dev, "Failed to allocate fbdev\n");
+ ret = -ENOMEM;
+ goto error_alloc_info;
+ }
+
+ host = to_imxfb_host(fb_info);
+
+ host->base = ioremap(res->start, resource_size(res));
+ if (!host->base) {
+ dev_err(&pdev->dev, "ioremap failed\n");
+ ret = -ENOMEM;
+ goto error_ioremap;
+ }
+
+ host->pdev = pdev;
+ platform_set_drvdata(pdev, host);
+
+ host->devdata = &mxsfb_devdata[pdev->id_entry->driver_data];
+
+ host->clk = clk_get(&host->pdev->dev, NULL);
+ if (IS_ERR(host->clk)) {
+ ret = PTR_ERR(host->clk);
+ goto error_getclock;
+ }
+
+ fb_info->pseudo_palette = kmalloc(sizeof(u32) * 16, GFP_KERNEL);
+ if (!fb_info->pseudo_palette) {
+ ret = -ENOMEM;
+ goto error_pseudo_pallette;
+ }
+
+ INIT_LIST_HEAD(&fb_info->modelist);
+
+ ret = mxsfb_init_fbinfo(host);
+ if (ret != 0)
+ goto error_init_fb;
+
+ for (i = 0; i < pdata->mode_count; i++)
+ fb_add_videomode(&pdata->mode_list[i], &fb_info->modelist);
+
+ modelist = list_first_entry(&fb_info->modelist,
+ struct fb_modelist, list);
+ fb_videomode_to_var(&fb_info->var, &modelist->mode);
+
+ /* init the color fields */
+ mxsfb_check_var(&fb_info->var, fb_info);
+
+ platform_set_drvdata(pdev, fb_info);
+
+ ret = register_framebuffer(fb_info);
+ if (ret != 0) {
+ dev_err(&pdev->dev,"Failed to register framebuffer\n");
+ goto error_register;
+ }
+
+ if (!host->enabled) {
+ writel(0, host->base + CTRL);
+ mxsfb_set_par(fb_info);
+ mxsfb_enable_controller(fb_info);
+ }
+
+ return 0;
+
+error_register:
+ if (host->enabled)
+ clk_disable(host->clk);
+ fb_destroy_modelist(&fb_info->modelist);
+error_init_fb:
+ kfree(fb_info->pseudo_palette);
+error_pseudo_pallette:
+ clk_put(host->clk);
+error_getclock:
+ iounmap(host->base);
+error_ioremap:
+ framebuffer_release(fb_info);
+error_alloc_info:
+ release_mem_region(res->start, resource_size(res));
+
+ return ret;
+}
+
+static int __devexit mxsfb_remove(struct platform_device *pdev)
+{
+ struct fb_info *fb_info = platform_get_drvdata(pdev);
+ struct mxsfb_info *host = to_imxfb_host(fb_info);
+ struct resource *res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+
+ mxsfb_disable_controller(fb_info);
+
+ unregister_framebuffer(fb_info);
+ kfree(fb_info->pseudo_palette);
+ mxsfb_free_videomem(host);
+ iounmap(host->base);
+ clk_put(host->clk);
+
+ framebuffer_release(fb_info);
+ release_mem_region(res->start, resource_size(res));
+
+ platform_set_drvdata(pdev, NULL);
+
+ return 0;
+}
+
+static struct platform_device_id mxsfb_devtype[] = {
+ {
+ .name = "imx23-fb",
+ .driver_data = MXSFB_MX23,
+ }, {
+ .name = "imx28-fb",
+ .driver_data = MXSFB_MX28,
+ }, {
+ },
+};
+MODULE_DEVICE_TABLE(platform, mxsfb_devtype);
+
+static struct platform_driver mxsfb_driver = {
+ .probe = mxsfb_probe,
+ .remove = __devexit_p(mxsfb_remove),
+ .id_table = mxsfb_devtype,
+ .driver = {
+ .name = DRIVER_NAME,
+ },
+};
+
+static int __init mxsfb_init(void)
+{
+ return platform_driver_register(&mxsfb_driver);
+}
+
+static void __exit mxsfb_exit(void)
+{
+ platform_driver_unregister(&mxsfb_driver);
+}
+
+module_init(mxsfb_init);
+module_exit(mxsfb_exit);
+
+MODULE_DESCRIPTION("Freescale mxs framebuffer driver");
+MODULE_AUTHOR("Sascha Hauer, Pengutronix");
+MODULE_LICENSE("GPL");
--
1.7.2.3
^ permalink raw reply related [flat|nested] 65+ messages in thread
* [PATCH 2/2] ARM i.MX23/28: Add framebuffer device support
2011-02-09 13:20 [PATCH] i.MX23/28 framebuffer driver Sascha Hauer
2011-02-09 13:20 ` Sascha Hauer
@ 2011-02-09 13:20 ` Sascha Hauer
2011-02-09 14:33 ` Fabio Estevam
2011-02-10 3:23 ` Li Frank-B20596
2011-02-09 14:47 ` [PATCH] i.MX23/28 framebuffer driver Arnd Bergmann
2 siblings, 2 replies; 65+ messages in thread
From: Sascha Hauer @ 2011-02-09 13:20 UTC (permalink / raw)
To: linux-arm-kernel
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Cc: Shawn Guo <shawn.guo@freescale.com>
---
arch/arm/mach-mxs/clock-mx23.c | 2 +-
arch/arm/mach-mxs/clock-mx28.c | 1 +
arch/arm/mach-mxs/devices-mx23.h | 4 ++
arch/arm/mach-mxs/devices-mx28.h | 4 ++
arch/arm/mach-mxs/devices/Kconfig | 4 ++
arch/arm/mach-mxs/devices/Makefile | 1 +
arch/arm/mach-mxs/devices/platform-mxsfb.c | 46 ++++++++++++++++++++++++++++
7 files changed, 61 insertions(+), 1 deletions(-)
create mode 100644 arch/arm/mach-mxs/devices/platform-mxsfb.c
diff --git a/arch/arm/mach-mxs/clock-mx23.c b/arch/arm/mach-mxs/clock-mx23.c
index ca72a05..bfc7f27 100644
--- a/arch/arm/mach-mxs/clock-mx23.c
+++ b/arch/arm/mach-mxs/clock-mx23.c
@@ -446,7 +446,7 @@ static struct clk_lookup lookups[] = {
_REGISTER_CLOCK(NULL, "hclk", hbus_clk)
_REGISTER_CLOCK(NULL, "usb", usb_clk)
_REGISTER_CLOCK(NULL, "audio", audio_clk)
- _REGISTER_CLOCK(NULL, "pwm", pwm_clk)
+ _REGISTER_CLOCK("imx23-fb", NULL, lcdif_clk)
};
static int clk_misc_init(void)
diff --git a/arch/arm/mach-mxs/clock-mx28.c b/arch/arm/mach-mxs/clock-mx28.c
index fd1c4c5..6a7ebcb 100644
--- a/arch/arm/mach-mxs/clock-mx28.c
+++ b/arch/arm/mach-mxs/clock-mx28.c
@@ -620,6 +620,7 @@ static struct clk_lookup lookups[] = {
_REGISTER_CLOCK(NULL, "pwm", pwm_clk)
_REGISTER_CLOCK(NULL, "lradc", lradc_clk)
_REGISTER_CLOCK(NULL, "spdif", spdif_clk)
+ _REGISTER_CLOCK("imx28-fb", NULL, lcdif_clk)
};
static int clk_misc_init(void)
diff --git a/arch/arm/mach-mxs/devices-mx23.h b/arch/arm/mach-mxs/devices-mx23.h
index 1256788..b9745a2 100644
--- a/arch/arm/mach-mxs/devices-mx23.h
+++ b/arch/arm/mach-mxs/devices-mx23.h
@@ -10,7 +10,11 @@
*/
#include <mach/mx23.h>
#include <mach/devices-common.h>
+#include <mach/fb.h>
extern const struct amba_device mx23_duart_device __initconst;
#define mx23_add_duart() \
mxs_add_duart(&mx23_duart_device)
+
+struct platform_device *__init mx23_add_mxsfb(
+ const struct mxsfb_platform_data *pdata);
diff --git a/arch/arm/mach-mxs/devices-mx28.h b/arch/arm/mach-mxs/devices-mx28.h
index 33773a6..c902dc7 100644
--- a/arch/arm/mach-mxs/devices-mx28.h
+++ b/arch/arm/mach-mxs/devices-mx28.h
@@ -10,6 +10,7 @@
*/
#include <mach/mx28.h>
#include <mach/devices-common.h>
+#include <mach/fb.h>
extern const struct amba_device mx28_duart_device __initconst;
#define mx28_add_duart() \
@@ -18,3 +19,6 @@ extern const struct amba_device mx28_duart_device __initconst;
extern const struct mxs_fec_data mx28_fec_data[] __initconst;
#define mx28_add_fec(id, pdata) \
mxs_add_fec(&mx28_fec_data[id], pdata)
+
+struct platform_device *__init mx28_add_mxsfb(
+ const struct mxsfb_platform_data *pdata);
diff --git a/arch/arm/mach-mxs/devices/Kconfig b/arch/arm/mach-mxs/devices/Kconfig
index cf7dc1a..1538cb9 100644
--- a/arch/arm/mach-mxs/devices/Kconfig
+++ b/arch/arm/mach-mxs/devices/Kconfig
@@ -4,3 +4,7 @@ config MXS_HAVE_AMBA_DUART
config MXS_HAVE_PLATFORM_FEC
bool
+
+config MXS_HAVE_PLATFORM_MXSFB
+ bool
+
diff --git a/arch/arm/mach-mxs/devices/Makefile b/arch/arm/mach-mxs/devices/Makefile
index d0a09f6..0cf8b48 100644
--- a/arch/arm/mach-mxs/devices/Makefile
+++ b/arch/arm/mach-mxs/devices/Makefile
@@ -1,2 +1,3 @@
obj-$(CONFIG_MXS_HAVE_AMBA_DUART) += amba-duart.o
obj-$(CONFIG_MXS_HAVE_PLATFORM_FEC) += platform-fec.o
+obj-$(CONFIG_MXS_HAVE_PLATFORM_MXSFB) += platform-mxsfb.o
diff --git a/arch/arm/mach-mxs/devices/platform-mxsfb.c b/arch/arm/mach-mxs/devices/platform-mxsfb.c
new file mode 100644
index 0000000..632bbdc
--- /dev/null
+++ b/arch/arm/mach-mxs/devices/platform-mxsfb.c
@@ -0,0 +1,46 @@
+/*
+ * Copyright (C) 2011 Pengutronix, Sascha Hauer <s.hauer@pengutronix.de>
+ *
+ * This program is free software; you can redistribute it and/or modify it under
+ * the terms of the GNU General Public License version 2 as published by the
+ * Free Software Foundation.
+ */
+#include <asm/sizes.h>
+#include <mach/mx23.h>
+#include <mach/mx28.h>
+#include <mach/devices-common.h>
+#include <mach/fb.h>
+
+#ifdef CONFIG_SOC_IMX23
+struct platform_device *__init mx23_add_mxsfb(
+ const struct mxsfb_platform_data *pdata)
+{
+ struct resource res[] = {
+ {
+ .start = MX23_LCDIF_BASE_ADDR,
+ .end = MX23_LCDIF_BASE_ADDR + SZ_8K - 1,
+ .flags = IORESOURCE_MEM,
+ },
+ };
+
+ return mxs_add_platform_device_dmamask("imx23-fb", -1,
+ res, ARRAY_SIZE(res), pdata, sizeof(*pdata), DMA_BIT_MASK(32));
+}
+#endif /* ifdef CONFIG_SOC_IMX23 */
+
+#ifdef CONFIG_SOC_IMX28
+struct platform_device *__init mx28_add_mxsfb(
+ const struct mxsfb_platform_data *pdata)
+{
+ struct resource res[] = {
+ {
+ .start = MX28_LCDIF_BASE_ADDR,
+ .end = MX28_LCDIF_BASE_ADDR + SZ_8K - 1,
+ .flags = IORESOURCE_MEM,
+ },
+ };
+
+ return mxs_add_platform_device_dmamask("imx28-fb", -1,
+ res, ARRAY_SIZE(res), pdata, sizeof(*pdata), DMA_BIT_MASK(32));
+}
+#endif /* ifdef CONFIG_SOC_IMX28 */
--
1.7.2.3
^ permalink raw reply related [flat|nested] 65+ messages in thread
* Re: [PATCH 1/2] video: Add i.MX23/28 framebuffer driver
2011-02-09 13:20 ` Sascha Hauer
@ 2011-02-09 13:35 ` Lothar Waßmann
-1 siblings, 0 replies; 65+ messages in thread
From: Lothar Waßmann @ 2011-02-09 13:35 UTC (permalink / raw)
To: linux-arm-kernel
Hi Sascha,
Sascha Hauer writes:
> Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
> Cc: Paul Mundt <lethal@linux-sh.org>
> Cc: linux-fbdev@vger.kernel.org
> Cc: Shawn Guo <shawn.guo@freescale.com>
> ---
> arch/arm/mach-mxs/include/mach/fb.h | 48 ++
> drivers/video/Kconfig | 9 +
> drivers/video/Makefile | 1 +
> drivers/video/mxsfb.c | 914 +++++++++++++++++++++++++++++++++++
> 4 files changed, 972 insertions(+), 0 deletions(-)
> create mode 100644 arch/arm/mach-mxs/include/mach/fb.h
> create mode 100644 drivers/video/mxsfb.c
>
[...]
> diff --git a/drivers/video/mxsfb.c b/drivers/video/mxsfb.c
> new file mode 100644
> index 0000000..92a5804
> --- /dev/null
> +++ b/drivers/video/mxsfb.c
[...]
> +static int __devexit mxsfb_remove(struct platform_device *pdev)
> +{
> + struct fb_info *fb_info = platform_get_drvdata(pdev);
> + struct mxsfb_info *host = to_imxfb_host(fb_info);
> + struct resource *res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
> +
> + mxsfb_disable_controller(fb_info);
>
IMO this should also be guarded by an:
| if (host->enabled)
otherwise it may lead to an unbalanced clk_disable() call.
Lothar Waßmann
--
___________________________________________________________
Ka-Ro electronics GmbH | Pascalstraße 22 | D - 52076 Aachen
Phone: +49 2408 1402-0 | Fax: +49 2408 1402-10
Geschäftsführer: Matthias Kaussen
Handelsregistereintrag: Amtsgericht Aachen, HRB 4996
www.karo-electronics.de | info@karo-electronics.de
___________________________________________________________
^ permalink raw reply [flat|nested] 65+ messages in thread
* [PATCH 1/2] video: Add i.MX23/28 framebuffer driver
@ 2011-02-09 13:35 ` Lothar Waßmann
0 siblings, 0 replies; 65+ messages in thread
From: Lothar Waßmann @ 2011-02-09 13:35 UTC (permalink / raw)
To: linux-arm-kernel
Hi Sascha,
Sascha Hauer writes:
> Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
> Cc: Paul Mundt <lethal@linux-sh.org>
> Cc: linux-fbdev at vger.kernel.org
> Cc: Shawn Guo <shawn.guo@freescale.com>
> ---
> arch/arm/mach-mxs/include/mach/fb.h | 48 ++
> drivers/video/Kconfig | 9 +
> drivers/video/Makefile | 1 +
> drivers/video/mxsfb.c | 914 +++++++++++++++++++++++++++++++++++
> 4 files changed, 972 insertions(+), 0 deletions(-)
> create mode 100644 arch/arm/mach-mxs/include/mach/fb.h
> create mode 100644 drivers/video/mxsfb.c
>
[...]
> diff --git a/drivers/video/mxsfb.c b/drivers/video/mxsfb.c
> new file mode 100644
> index 0000000..92a5804
> --- /dev/null
> +++ b/drivers/video/mxsfb.c
[...]
> +static int __devexit mxsfb_remove(struct platform_device *pdev)
> +{
> + struct fb_info *fb_info = platform_get_drvdata(pdev);
> + struct mxsfb_info *host = to_imxfb_host(fb_info);
> + struct resource *res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
> +
> + mxsfb_disable_controller(fb_info);
>
IMO this should also be guarded by an:
| if (host->enabled)
otherwise it may lead to an unbalanced clk_disable() call.
Lothar Wa?mann
--
___________________________________________________________
Ka-Ro electronics GmbH | Pascalstra?e 22 | D - 52076 Aachen
Phone: +49 2408 1402-0 | Fax: +49 2408 1402-10
Gesch?ftsf?hrer: Matthias Kaussen
Handelsregistereintrag: Amtsgericht Aachen, HRB 4996
www.karo-electronics.de | info at karo-electronics.de
___________________________________________________________
^ permalink raw reply [flat|nested] 65+ messages in thread
* [PATCH 2/2] ARM i.MX23/28: Add framebuffer device support
2011-02-09 13:20 ` [PATCH 2/2] ARM i.MX23/28: Add framebuffer device support Sascha Hauer
@ 2011-02-09 14:33 ` Fabio Estevam
2011-02-10 3:23 ` Li Frank-B20596
1 sibling, 0 replies; 65+ messages in thread
From: Fabio Estevam @ 2011-02-09 14:33 UTC (permalink / raw)
To: linux-arm-kernel
Hi Sascha,
On Wed, Feb 9, 2011 at 11:20 AM, Sascha Hauer <s.hauer@pengutronix.de> wrote:
> Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
> Cc: Shawn Guo <shawn.guo@freescale.com>
> ---
> ?arch/arm/mach-mxs/clock-mx23.c ? ? ? ? ? ? | ? ?2 +-
> ?arch/arm/mach-mxs/clock-mx28.c ? ? ? ? ? ? | ? ?1 +
> ?arch/arm/mach-mxs/devices-mx23.h ? ? ? ? ? | ? ?4 ++
> ?arch/arm/mach-mxs/devices-mx28.h ? ? ? ? ? | ? ?4 ++
> ?arch/arm/mach-mxs/devices/Kconfig ? ? ? ? ?| ? ?4 ++
> ?arch/arm/mach-mxs/devices/Makefile ? ? ? ? | ? ?1 +
> ?arch/arm/mach-mxs/devices/platform-mxsfb.c | ? 46 ++++++++++++++++++++++++++++
> ?7 files changed, 61 insertions(+), 1 deletions(-)
> ?create mode 100644 arch/arm/mach-mxs/devices/platform-mxsfb.c
>
> diff --git a/arch/arm/mach-mxs/clock-mx23.c b/arch/arm/mach-mxs/clock-mx23.c
> index ca72a05..bfc7f27 100644
> --- a/arch/arm/mach-mxs/clock-mx23.c
> +++ b/arch/arm/mach-mxs/clock-mx23.c
> @@ -446,7 +446,7 @@ static struct clk_lookup lookups[] = {
> ? ? ? ?_REGISTER_CLOCK(NULL, "hclk", hbus_clk)
> ? ? ? ?_REGISTER_CLOCK(NULL, "usb", usb_clk)
> ? ? ? ?_REGISTER_CLOCK(NULL, "audio", audio_clk)
> - ? ? ? _REGISTER_CLOCK(NULL, "pwm", pwm_clk)
> + ? ? ? _REGISTER_CLOCK("imx23-fb", NULL, lcdif_clk)
Is the removal of the pwm clock intended here?
Regards,
Fabio Estevam
^ permalink raw reply [flat|nested] 65+ messages in thread
* [PATCH] i.MX23/28 framebuffer driver
2011-02-09 13:20 [PATCH] i.MX23/28 framebuffer driver Sascha Hauer
2011-02-09 13:20 ` Sascha Hauer
2011-02-09 13:20 ` [PATCH 2/2] ARM i.MX23/28: Add framebuffer device support Sascha Hauer
@ 2011-02-09 14:47 ` Arnd Bergmann
2011-02-09 15:37 ` Sascha Hauer
2 siblings, 1 reply; 65+ messages in thread
From: Arnd Bergmann @ 2011-02-09 14:47 UTC (permalink / raw)
To: linux-arm-kernel
On Wednesday 09 February 2011, Sascha Hauer wrote:
> Hi all,
>
> The following patches add framebuffer support for Freescale i.MX23/28
> processors.
Hi Sascha,
The driver looks good as far as I can tell, but I'd recommend
submitting it to linux-fbdev at vger.kernel.org for review in
addition to linux-arm-kernel. The frame buffer layer is largely
unmaintained these days, but there are still people on that
list that care.
Did you consider making the driver a KMS driver instead of
a frame buffer? I think the recommendation these days is
to start out with KMS for new drivers, which will be somewhat
simpler and give you a frame buffer device as well. I don't
think that there is a need to change over any drivers from
fb to kms though, since you've already done the work.
Arnd
^ permalink raw reply [flat|nested] 65+ messages in thread
* [PATCH] i.MX23/28 framebuffer driver
2011-02-09 14:47 ` [PATCH] i.MX23/28 framebuffer driver Arnd Bergmann
@ 2011-02-09 15:37 ` Sascha Hauer
2011-02-09 16:31 ` Arnd Bergmann
2011-02-10 9:22 ` Domenico Andreoli
0 siblings, 2 replies; 65+ messages in thread
From: Sascha Hauer @ 2011-02-09 15:37 UTC (permalink / raw)
To: linux-arm-kernel
Hi Arnd,
On Wed, Feb 09, 2011 at 03:47:11PM +0100, Arnd Bergmann wrote:
> On Wednesday 09 February 2011, Sascha Hauer wrote:
> > Hi all,
> >
> > The following patches add framebuffer support for Freescale i.MX23/28
> > processors.
>
> Hi Sascha,
>
> The driver looks good as far as I can tell, but I'd recommend
> submitting it to linux-fbdev at vger.kernel.org for review in
> addition to linux-arm-kernel. The frame buffer layer is largely
> unmaintained these days, but there are still people on that
> list that care.
The driver patch itself is Cced to linux-fbdev, only the introductory
mail is not.
>
> Did you consider making the driver a KMS driver instead of
> a frame buffer? I think the recommendation these days is
> to start out with KMS for new drivers, which will be somewhat
> simpler and give you a frame buffer device as well. I don't
> think that there is a need to change over any drivers from
> fb to kms though, since you've already done the work.
I tried doing so for the i.MX51 which supports multiple displays on dvi
and vga outputs and thus could make good use of KMS and friends. Anyway,
I got stuck quite fast. The KMS stuff is tightly coupled with DRM/DRI
and needs many many callbacks to implement. Additionally the userspace
tools expect a nvidia/amd/intel driver and do not have a generic
fallback. I think this stuff is good for implementing a full blown
graphics driver, but is lacking support for simple framebuffer grapics.
I'd love to go this way but it still requires a lot of work.
Sascha
--
Pengutronix e.K. | |
Industrial Linux Solutions | http://www.pengutronix.de/ |
Peiner Str. 6-8, 31137 Hildesheim, Germany | Phone: +49-5121-206917-0 |
Amtsgericht Hildesheim, HRA 2686 | Fax: +49-5121-206917-5555 |
^ permalink raw reply [flat|nested] 65+ messages in thread
* Re: [PATCH] i.MX23/28 framebuffer driver
2011-02-09 15:37 ` Sascha Hauer
@ 2011-02-09 16:31 ` Arnd Bergmann
2011-02-10 9:22 ` Domenico Andreoli
1 sibling, 0 replies; 65+ messages in thread
From: Arnd Bergmann @ 2011-02-09 16:31 UTC (permalink / raw)
To: linux-arm-kernel
On Wednesday 09 February 2011, Sascha Hauer wrote:
> The driver patch itself is Cced to linux-fbdev, only the introductory
> mail is not.
Ok, I see.
> > Did you consider making the driver a KMS driver instead of
> > a frame buffer? I think the recommendation these days is
> > to start out with KMS for new drivers, which will be somewhat
> > simpler and give you a frame buffer device as well. I don't
> > think that there is a need to change over any drivers from
> > fb to kms though, since you've already done the work.
>
> I tried doing so for the i.MX51 which supports multiple displays on dvi
> and vga outputs and thus could make good use of KMS and friends. Anyway,
> I got stuck quite fast. The KMS stuff is tightly coupled with DRM/DRI
> and needs many many callbacks to implement. Additionally the userspace
> tools expect a nvidia/amd/intel driver and do not have a generic
> fallback. I think this stuff is good for implementing a full blown
> graphics driver, but is lacking support for simple framebuffer grapics.
> I'd love to go this way but it still requires a lot of work.
Ok. This sounds like a lot of upfront work indeed, to make KMS more
generic, though I think a number of driver would benefit from it
eventually. It could be something for the Linaro graphics working
group to look at in the following 11.11 release, depending on how
many other people are interested in getting there.
Arnd
^ permalink raw reply [flat|nested] 65+ messages in thread
* [PATCH] i.MX23/28 framebuffer driver
@ 2011-02-09 16:31 ` Arnd Bergmann
0 siblings, 0 replies; 65+ messages in thread
From: Arnd Bergmann @ 2011-02-09 16:31 UTC (permalink / raw)
To: linux-arm-kernel
On Wednesday 09 February 2011, Sascha Hauer wrote:
> The driver patch itself is Cced to linux-fbdev, only the introductory
> mail is not.
Ok, I see.
> > Did you consider making the driver a KMS driver instead of
> > a frame buffer? I think the recommendation these days is
> > to start out with KMS for new drivers, which will be somewhat
> > simpler and give you a frame buffer device as well. I don't
> > think that there is a need to change over any drivers from
> > fb to kms though, since you've already done the work.
>
> I tried doing so for the i.MX51 which supports multiple displays on dvi
> and vga outputs and thus could make good use of KMS and friends. Anyway,
> I got stuck quite fast. The KMS stuff is tightly coupled with DRM/DRI
> and needs many many callbacks to implement. Additionally the userspace
> tools expect a nvidia/amd/intel driver and do not have a generic
> fallback. I think this stuff is good for implementing a full blown
> graphics driver, but is lacking support for simple framebuffer grapics.
> I'd love to go this way but it still requires a lot of work.
Ok. This sounds like a lot of upfront work indeed, to make KMS more
generic, though I think a number of driver would benefit from it
eventually. It could be something for the Linaro graphics working
group to look at in the following 11.11 release, depending on how
many other people are interested in getting there.
Arnd
^ permalink raw reply [flat|nested] 65+ messages in thread
* RE: [PATCH 1/2] video: Add i.MX23/28 framebuffer driver
2011-02-09 13:20 ` Sascha Hauer
@ 2011-02-10 3:16 ` Li Frank-B20596
-1 siblings, 0 replies; 65+ messages in thread
From: Li Frank-B20596 @ 2011-02-10 3:16 UTC (permalink / raw)
To: linux-arm-kernel
> +#define CTRL 0x00
> +#define CTRL1 0x10
> +#define MX28_CTRL2 0x20
> +#define MX23_TRANSFER_COUNT 0x20
> +#define MX28_TRANSFER_COUNT 0x30
> +#define MX28_CUR_BUF 0x40
> +#define MX28_NEXT_BUF 0x50
> +#define MX23_CUR_BUF 0x30
> +#define MX23_NEXT_BUF 0x40
> +#define TIMING 0x60
> +#define VDCTRL0 0x70
> +#define VDCTRL1 0x80
> +#define VDCTRL2 0x90
> +#define VDCTRL3 0xa0
> +#define VDCTRL4 0xb0
Why you give up mx23/mx28 register define role, which generate from SOC xml.
There is a set header files for each mx23/mx28 module, which generate from xml.
I know original header files affect run time one Image.
But I think we can copy common part of such register definition because
That keep consistent with mx23/mx28 data sheet. Data sheet and header file generate
from one source xml.
HW_<Module name>_<Register name>.
BM_<Module name>_<Register name>_Bit name.
^ permalink raw reply [flat|nested] 65+ messages in thread
* [PATCH 1/2] video: Add i.MX23/28 framebuffer driver
@ 2011-02-10 3:16 ` Li Frank-B20596
0 siblings, 0 replies; 65+ messages in thread
From: Li Frank-B20596 @ 2011-02-10 3:16 UTC (permalink / raw)
To: linux-arm-kernel
> +#define CTRL 0x00
> +#define CTRL1 0x10
> +#define MX28_CTRL2 0x20
> +#define MX23_TRANSFER_COUNT 0x20
> +#define MX28_TRANSFER_COUNT 0x30
> +#define MX28_CUR_BUF 0x40
> +#define MX28_NEXT_BUF 0x50
> +#define MX23_CUR_BUF 0x30
> +#define MX23_NEXT_BUF 0x40
> +#define TIMING 0x60
> +#define VDCTRL0 0x70
> +#define VDCTRL1 0x80
> +#define VDCTRL2 0x90
> +#define VDCTRL3 0xa0
> +#define VDCTRL4 0xb0
Why you give up mx23/mx28 register define role, which generate from SOC xml.
There is a set header files for each mx23/mx28 module, which generate from xml.
I know original header files affect run time one Image.
But I think we can copy common part of such register definition because
That keep consistent with mx23/mx28 data sheet. Data sheet and header file generate
from one source xml.
HW_<Module name>_<Register name>.
BM_<Module name>_<Register name>_Bit name.
^ permalink raw reply [flat|nested] 65+ messages in thread
* [PATCH 2/2] ARM i.MX23/28: Add framebuffer device support
2011-02-09 13:20 ` [PATCH 2/2] ARM i.MX23/28: Add framebuffer device support Sascha Hauer
2011-02-09 14:33 ` Fabio Estevam
@ 2011-02-10 3:23 ` Li Frank-B20596
1 sibling, 0 replies; 65+ messages in thread
From: Li Frank-B20596 @ 2011-02-10 3:23 UTC (permalink / raw)
To: linux-arm-kernel
> diff --git a/arch/arm/mach-mxs/clock-mx23.c b/arch/arm/mach-mxs/clock-
> mx23.c
> index ca72a05..bfc7f27 100644
> --- a/arch/arm/mach-mxs/clock-mx23.c
> +++ b/arch/arm/mach-mxs/clock-mx23.c
> @@ -446,7 +446,7 @@ static struct clk_lookup lookups[] = {
> _REGISTER_CLOCK(NULL, "hclk", hbus_clk)
> _REGISTER_CLOCK(NULL, "usb", usb_clk)
> _REGISTER_CLOCK(NULL, "audio", audio_clk)
> - _REGISTER_CLOCK(NULL, "pwm", pwm_clk)
> + _REGISTER_CLOCK("imx23-fb", NULL, lcdif_clk)
> };
Why remove mx23 pwm clk here?
^ permalink raw reply [flat|nested] 65+ messages in thread
* Re: [PATCH 1/2] video: Add i.MX23/28 framebuffer driver
2011-02-10 3:16 ` Li Frank-B20596
@ 2011-02-10 8:51 ` Juergen Beisert
-1 siblings, 0 replies; 65+ messages in thread
From: Juergen Beisert @ 2011-02-10 8:51 UTC (permalink / raw)
To: linux-arm-kernel
Li Frank-B20596 wrote:
> > +#define CTRL 0x00
> > +#define CTRL1 0x10
> > +#define MX28_CTRL2 0x20
> > +#define MX23_TRANSFER_COUNT 0x20
> > +#define MX28_TRANSFER_COUNT 0x30
> > +#define MX28_CUR_BUF 0x40
> > +#define MX28_NEXT_BUF 0x50
> > +#define MX23_CUR_BUF 0x30
> > +#define MX23_NEXT_BUF 0x40
> > +#define TIMING 0x60
> > +#define VDCTRL0 0x70
> > +#define VDCTRL1 0x80
> > +#define VDCTRL2 0x90
> > +#define VDCTRL3 0xa0
> > +#define VDCTRL4 0xb0
>
> Why you give up mx23/mx28 register define role, which generate from SOC
> xml.
Your macros prevent me from writing short and compact code. If you need more
than one of these macros you always have to split each line to follow the 80
columns rule. Unreadable.
> There is a set header files for each mx23/mx28 module, which generate
> from xml. I know original header files affect run time one Image.
> But I think we can copy common part of such register definition because
> That keep consistent with mx23/mx28 data sheet. Data sheet and header file
> generate from one source xml.
>
> HW_<Module name>_<Register name>.
> BM_<Module name>_<Register name>_Bit name.
IMHO when I define the macros where they belong to, there is not need for this
redundant HW_<Module name> or BW__<Module name> prefixes. They are just
needless.
Regards,
Juergen
--
Pengutronix e.K. | Juergen Beisert |
Linux Solutions for Science and Industry | Phone: +49-8766-939 228 |
Vertretung Sued/Muenchen, Germany | Fax: +49-5121-206917-5555 |
Amtsgericht Hildesheim, HRA 2686 | http://www.pengutronix.de/ |
^ permalink raw reply [flat|nested] 65+ messages in thread
* [PATCH 1/2] video: Add i.MX23/28 framebuffer driver
@ 2011-02-10 8:51 ` Juergen Beisert
0 siblings, 0 replies; 65+ messages in thread
From: Juergen Beisert @ 2011-02-10 8:51 UTC (permalink / raw)
To: linux-arm-kernel
Li Frank-B20596 wrote:
> > +#define CTRL 0x00
> > +#define CTRL1 0x10
> > +#define MX28_CTRL2 0x20
> > +#define MX23_TRANSFER_COUNT 0x20
> > +#define MX28_TRANSFER_COUNT 0x30
> > +#define MX28_CUR_BUF 0x40
> > +#define MX28_NEXT_BUF 0x50
> > +#define MX23_CUR_BUF 0x30
> > +#define MX23_NEXT_BUF 0x40
> > +#define TIMING 0x60
> > +#define VDCTRL0 0x70
> > +#define VDCTRL1 0x80
> > +#define VDCTRL2 0x90
> > +#define VDCTRL3 0xa0
> > +#define VDCTRL4 0xb0
>
> Why you give up mx23/mx28 register define role, which generate from SOC
> xml.
Your macros prevent me from writing short and compact code. If you need more
than one of these macros you always have to split each line to follow the 80
columns rule. Unreadable.
> There is a set header files for each mx23/mx28 module, which generate
> from xml. I know original header files affect run time one Image.
> But I think we can copy common part of such register definition because
> That keep consistent with mx23/mx28 data sheet. Data sheet and header file
> generate from one source xml.
>
> HW_<Module name>_<Register name>.
> BM_<Module name>_<Register name>_Bit name.
IMHO when I define the macros where they belong to, there is not need for this
redundant HW_<Module name> or BW__<Module name> prefixes. They are just
needless.
Regards,
Juergen
--
Pengutronix e.K. | Juergen Beisert |
Linux Solutions for Science and Industry | Phone: +49-8766-939 228 |
Vertretung Sued/Muenchen, Germany | Fax: +49-5121-206917-5555 |
Amtsgericht Hildesheim, HRA 2686 | http://www.pengutronix.de/ |
^ permalink raw reply [flat|nested] 65+ messages in thread
* RE: [PATCH 1/2] video: Add i.MX23/28 framebuffer driver
2011-02-10 8:51 ` Juergen Beisert
@ 2011-02-10 9:15 ` Li Frank-B20596
-1 siblings, 0 replies; 65+ messages in thread
From: Li Frank-B20596 @ 2011-02-10 9:15 UTC (permalink / raw)
To: linux-arm-kernel
> > > +#define VDCTRL1 0x80
> > > +#define VDCTRL2 0x90
> > > +#define VDCTRL3 0xa0
> > > +#define VDCTRL4 0xb0
> >
> > Why you give up mx23/mx28 register define role, which generate from SOC
> > xml.
>
> Your macros prevent me from writing short and compact code. If you need
> more
> than one of these macros you always have to split each line to follow the
> 80
> columns rule. Unreadable.
>
> > There is a set header files for each mx23/mx28 module, which generate
> > from xml. I know original header files affect run time one Image.
> > But I think we can copy common part of such register definition because
> > That keep consistent with mx23/mx28 data sheet. Data sheet and header
> file
> > generate from one source xml.
> >
> > HW_<Module name>_<Register name>.
> > BM_<Module name>_<Register name>_Bit name.
>
> IMHO when I define the macros where they belong to, there is not need for
> this
> redundant HW_<Module name> or BW__<Module name> prefixes. They are just
> needless.
At first, someone complain name is longer. But during mx23/mx28 developing,
Everyone start enjoy such definition because there are not error happen about
register and bit position definition and identical map to silicon spec.
Developer needn't look up register header file when coding, just write down
Register name or bit name according to mx23/mx28 spec.
If you still think it is too long, I suggest keep HW_ and BM_ prefix to distinguish
Which one is register name, which one is bit mask.
Imx23/imx28 register have consistent convention.
HW_ is register name
BM_ is bit mask
BP_ is bit position.
BF_(x) is ((x<<BP)&BM_)
>
> Regards,
> Juergen
>
^ permalink raw reply [flat|nested] 65+ messages in thread
* [PATCH 1/2] video: Add i.MX23/28 framebuffer driver
@ 2011-02-10 9:15 ` Li Frank-B20596
0 siblings, 0 replies; 65+ messages in thread
From: Li Frank-B20596 @ 2011-02-10 9:15 UTC (permalink / raw)
To: linux-arm-kernel
> > > +#define VDCTRL1 0x80
> > > +#define VDCTRL2 0x90
> > > +#define VDCTRL3 0xa0
> > > +#define VDCTRL4 0xb0
> >
> > Why you give up mx23/mx28 register define role, which generate from SOC
> > xml.
>
> Your macros prevent me from writing short and compact code. If you need
> more
> than one of these macros you always have to split each line to follow the
> 80
> columns rule. Unreadable.
>
> > There is a set header files for each mx23/mx28 module, which generate
> > from xml. I know original header files affect run time one Image.
> > But I think we can copy common part of such register definition because
> > That keep consistent with mx23/mx28 data sheet. Data sheet and header
> file
> > generate from one source xml.
> >
> > HW_<Module name>_<Register name>.
> > BM_<Module name>_<Register name>_Bit name.
>
> IMHO when I define the macros where they belong to, there is not need for
> this
> redundant HW_<Module name> or BW__<Module name> prefixes. They are just
> needless.
At first, someone complain name is longer. But during mx23/mx28 developing,
Everyone start enjoy such definition because there are not error happen about
register and bit position definition and identical map to silicon spec.
Developer needn't look up register header file when coding, just write down
Register name or bit name according to mx23/mx28 spec.
If you still think it is too long, I suggest keep HW_ and BM_ prefix to distinguish
Which one is register name, which one is bit mask.
Imx23/imx28 register have consistent convention.
HW_ is register name
BM_ is bit mask
BP_ is bit position.
BF_(x) is ((x<<BP)&BM_)
>
> Regards,
> Juergen
>
^ permalink raw reply [flat|nested] 65+ messages in thread
* [PATCH] i.MX23/28 framebuffer driver
2011-02-09 15:37 ` Sascha Hauer
2011-02-09 16:31 ` Arnd Bergmann
@ 2011-02-10 9:22 ` Domenico Andreoli
1 sibling, 0 replies; 65+ messages in thread
From: Domenico Andreoli @ 2011-02-10 9:22 UTC (permalink / raw)
To: linux-arm-kernel
Hi Sascha,
On Wed, Feb 09, 2011 at 04:37:16PM +0100, ext Sascha Hauer wrote:
> Hi Arnd,
>
> On Wed, Feb 09, 2011 at 03:47:11PM +0100, Arnd Bergmann wrote:
> >
> > Did you consider making the driver a KMS driver instead of
> > a frame buffer? I think the recommendation these days is
> > to start out with KMS for new drivers, which will be somewhat
> > simpler and give you a frame buffer device as well. I don't
> > think that there is a need to change over any drivers from
> > fb to kms though, since you've already done the work.
>
> I tried doing so for the i.MX51 which supports multiple displays on dvi
> and vga outputs and thus could make good use of KMS and friends. Anyway,
> I got stuck quite fast. The KMS stuff is tightly coupled with DRM/DRI
> and needs many many callbacks to implement. Additionally the userspace
> tools expect a nvidia/amd/intel driver and do not have a generic
> fallback. I think this stuff is good for implementing a full blown
> graphics driver, but is lacking support for simple framebuffer grapics.
> I'd love to go this way but it still requires a lot of work.
I will get the i.MX28 EVK within a couple of weeks. I'm curious of KMS,
which I don't know at all, and I would like to work on it. We could
join the efforts. What do you think?
Best regards,
Domenico
^ permalink raw reply [flat|nested] 65+ messages in thread
* RE: [PATCH 1/2] video: Add i.MX23/28 framebuffer driver
2011-02-09 13:20 ` Sascha Hauer
@ 2011-02-10 9:23 ` Li Frank-B20596
-1 siblings, 0 replies; 65+ messages in thread
From: Li Frank-B20596 @ 2011-02-10 9:23 UTC (permalink / raw)
To: linux-arm-kernel
> +
> + /* if it was disabled, re-enable the mode again */
> + reg = readl(host->base + CTRL);
> + reg |= CTRL_DOTCLK_MODE;
> + writel(reg, host->base + CTRL);
writel(CTRL_DOTCLK_MODE, host->base + CTRL_SET)
CTRL_SET is CTRL+0x4
SET and CLR register is easier than Read and write back.
^ permalink raw reply [flat|nested] 65+ messages in thread
* [PATCH 1/2] video: Add i.MX23/28 framebuffer driver
@ 2011-02-10 9:23 ` Li Frank-B20596
0 siblings, 0 replies; 65+ messages in thread
From: Li Frank-B20596 @ 2011-02-10 9:23 UTC (permalink / raw)
To: linux-arm-kernel
> +
> + /* if it was disabled, re-enable the mode again */
> + reg = readl(host->base + CTRL);
> + reg |= CTRL_DOTCLK_MODE;
> + writel(reg, host->base + CTRL);
writel(CTRL_DOTCLK_MODE, host->base + CTRL_SET)
CTRL_SET is CTRL+0x4
SET and CLR register is easier than Read and write back.
^ permalink raw reply [flat|nested] 65+ messages in thread
* Re: [PATCH 1/2] video: Add i.MX23/28 framebuffer driver
2011-02-10 9:23 ` Li Frank-B20596
@ 2011-02-10 9:46 ` Juergen Beisert
-1 siblings, 0 replies; 65+ messages in thread
From: Juergen Beisert @ 2011-02-10 9:46 UTC (permalink / raw)
To: linux-arm-kernel
Li Frank-B20596 wrote:
> > +
> > + /* if it was disabled, re-enable the mode again */
> > + reg = readl(host->base + CTRL);
> > + reg |= CTRL_DOTCLK_MODE;
> > + writel(reg, host->base + CTRL);
>
> writel(CTRL_DOTCLK_MODE, host->base + CTRL_SET)
> CTRL_SET is CTRL+0x4
> SET and CLR register is easier than Read and write back.
But you must always check, if the register really has such a SET and CLEAR
feature. Not all registers have this feature...
But in this case you are right.
But I would prefer:
#define SET 4
#define CLEAR 8
[...]
writel(CTRL_DOTCLK_MODE, host->base + CTRL + SET)
:-)
Regards,
Juergen
--
Pengutronix e.K. | Juergen Beisert |
Linux Solutions for Science and Industry | Phone: +49-8766-939 228 |
Vertretung Sued/Muenchen, Germany | Fax: +49-5121-206917-5555 |
Amtsgericht Hildesheim, HRA 2686 | http://www.pengutronix.de/ |
^ permalink raw reply [flat|nested] 65+ messages in thread
* [PATCH 1/2] video: Add i.MX23/28 framebuffer driver
@ 2011-02-10 9:46 ` Juergen Beisert
0 siblings, 0 replies; 65+ messages in thread
From: Juergen Beisert @ 2011-02-10 9:46 UTC (permalink / raw)
To: linux-arm-kernel
Li Frank-B20596 wrote:
> > +
> > + /* if it was disabled, re-enable the mode again */
> > + reg = readl(host->base + CTRL);
> > + reg |= CTRL_DOTCLK_MODE;
> > + writel(reg, host->base + CTRL);
>
> writel(CTRL_DOTCLK_MODE, host->base + CTRL_SET)
> CTRL_SET is CTRL+0x4
> SET and CLR register is easier than Read and write back.
But you must always check, if the register really has such a SET and CLEAR
feature. Not all registers have this feature...
But in this case you are right.
But I would prefer:
#define SET 4
#define CLEAR 8
[...]
writel(CTRL_DOTCLK_MODE, host->base + CTRL + SET)
:-)
Regards,
Juergen
--
Pengutronix e.K. | Juergen Beisert |
Linux Solutions for Science and Industry | Phone: +49-8766-939 228 |
Vertretung Sued/Muenchen, Germany | Fax: +49-5121-206917-5555 |
Amtsgericht Hildesheim, HRA 2686 | http://www.pengutronix.de/ |
^ permalink raw reply [flat|nested] 65+ messages in thread
* Re: [PATCH 1/2] video: Add i.MX23/28 framebuffer driver
2011-02-10 9:15 ` Li Frank-B20596
@ 2011-02-10 9:52 ` Juergen Beisert
-1 siblings, 0 replies; 65+ messages in thread
From: Juergen Beisert @ 2011-02-10 9:52 UTC (permalink / raw)
To: linux-arm-kernel
Li Frank-B20596 wrote:
> > > > +#define VDCTRL1 0x80
> > > > +#define VDCTRL2 0x90
> > > > +#define VDCTRL3 0xa0
> > > > +#define VDCTRL4 0xb0
> > >
> > > Why you give up mx23/mx28 register define role, which generate from SOC
> > > xml.
> >
> > Your macros prevent me from writing short and compact code. If you need
> > more than one of these macros you always have to split each line to follow
> > the 80 columns rule. Unreadable.
> >
> > > There is a set header files for each mx23/mx28 module, which generate
> > > from xml. I know original header files affect run time one Image.
> > > But I think we can copy common part of such register definition because
> > > That keep consistent with mx23/mx28 data sheet. Data sheet and header
> > > file
> >
> > > generate from one source xml.
> > >
> > > HW_<Module name>_<Register name>.
> > > BM_<Module name>_<Register name>_Bit name.
> >
> > IMHO when I define the macros where they belong to, there is not need for
> > this redundant HW_<Module name> or BW__<Module name> prefixes. They are
> > just needless.
>
> At first, someone complain name is longer. But during mx23/mx28 developing,
> Everyone start enjoy such definition because there are not error happen
> about register and bit position definition and identical map to silicon
> spec.
This kind of macro encryption _may_ help when you are coding the driver the
first time. But after reading and reading it again (while testing and
debugging) all these prefixes and suffixes do not add any valuable
information. They only fill up your lines, enlarges your source code and make
you blind for the real bug...
Everyone who wants to see how source code looks that uses these longs macros I
can recommend reading the so called 'bootlets' source code. :-))
> Developer needn't look up register header file when coding, just
That's why I add these macros into the source file instead into a header file.
> write down Register name or bit name according to mx23/mx28 spec.
And sometimes the spec is incomplete or just wrong and so on. Independent of
any macro naming policy.
> If you still think it is too long, I suggest keep HW_ and BM_ prefix to
> distinguish Which one is register name, which one is bit mask.
The used macro in the address part of the writel() must be an offset, while
the macro used in the value part must be a bit definition. Anything else is
redundant.
Regards,
Juergen
--
Pengutronix e.K. | Juergen Beisert |
Linux Solutions for Science and Industry | Phone: +49-8766-939 228 |
Vertretung Sued/Muenchen, Germany | Fax: +49-5121-206917-5555 |
Amtsgericht Hildesheim, HRA 2686 | http://www.pengutronix.de/ |
^ permalink raw reply [flat|nested] 65+ messages in thread
* [PATCH 1/2] video: Add i.MX23/28 framebuffer driver
@ 2011-02-10 9:52 ` Juergen Beisert
0 siblings, 0 replies; 65+ messages in thread
From: Juergen Beisert @ 2011-02-10 9:52 UTC (permalink / raw)
To: linux-arm-kernel
Li Frank-B20596 wrote:
> > > > +#define VDCTRL1 0x80
> > > > +#define VDCTRL2 0x90
> > > > +#define VDCTRL3 0xa0
> > > > +#define VDCTRL4 0xb0
> > >
> > > Why you give up mx23/mx28 register define role, which generate from SOC
> > > xml.
> >
> > Your macros prevent me from writing short and compact code. If you need
> > more than one of these macros you always have to split each line to follow
> > the 80 columns rule. Unreadable.
> >
> > > There is a set header files for each mx23/mx28 module, which generate
> > > from xml. I know original header files affect run time one Image.
> > > But I think we can copy common part of such register definition because
> > > That keep consistent with mx23/mx28 data sheet. Data sheet and header
> > > file
> >
> > > generate from one source xml.
> > >
> > > HW_<Module name>_<Register name>.
> > > BM_<Module name>_<Register name>_Bit name.
> >
> > IMHO when I define the macros where they belong to, there is not need for
> > this redundant HW_<Module name> or BW__<Module name> prefixes. They are
> > just needless.
>
> At first, someone complain name is longer. But during mx23/mx28 developing,
> Everyone start enjoy such definition because there are not error happen
> about register and bit position definition and identical map to silicon
> spec.
This kind of macro encryption _may_ help when you are coding the driver the
first time. But after reading and reading it again (while testing and
debugging) all these prefixes and suffixes do not add any valuable
information. They only fill up your lines, enlarges your source code and make
you blind for the real bug...
Everyone who wants to see how source code looks that uses these longs macros I
can recommend reading the so called 'bootlets' source code. :-))
> Developer needn't look up register header file when coding, just
That's why I add these macros into the source file instead into a header file.
> write down Register name or bit name according to mx23/mx28 spec.
And sometimes the spec is incomplete or just wrong and so on. Independent of
any macro naming policy.
> If you still think it is too long, I suggest keep HW_ and BM_ prefix to
> distinguish Which one is register name, which one is bit mask.
The used macro in the address part of the writel() must be an offset, while
the macro used in the value part must be a bit definition. Anything else is
redundant.
Regards,
Juergen
--
Pengutronix e.K. | Juergen Beisert |
Linux Solutions for Science and Industry | Phone: +49-8766-939 228 |
Vertretung Sued/Muenchen, Germany | Fax: +49-5121-206917-5555 |
Amtsgericht Hildesheim, HRA 2686 | http://www.pengutronix.de/ |
^ permalink raw reply [flat|nested] 65+ messages in thread
* RE: [PATCH 1/2] video: Add i.MX23/28 framebuffer driver
2011-02-10 9:46 ` Juergen Beisert
@ 2011-02-10 10:08 ` Li Frank-B20596
-1 siblings, 0 replies; 65+ messages in thread
From: Li Frank-B20596 @ 2011-02-10 10:08 UTC (permalink / raw)
To: linux-arm-kernel
> Li Frank-B20596 wrote:
> > > +
> > > + /* if it was disabled, re-enable the mode again */
> > > + reg = readl(host->base + CTRL);
> > > + reg |= CTRL_DOTCLK_MODE;
> > > + writel(reg, host->base + CTRL);
> >
> > writel(CTRL_DOTCLK_MODE, host->base + CTRL_SET)
> > CTRL_SET is CTRL+0x4
> > SET and CLR register is easier than Read and write back.
>
> But you must always check, if the register really has such a SET and
> CLEAR
> feature. Not all registers have this feature...
That's a reason why I suggest use register header that generate from xml.
HW_XX_CTRL_SET will exist if there are such register in hardware.
If hardware have not such register, there are not _SET and _CLR.
Compiler will report error if user use _SET but hardware not such register.
>
> But in this case you are right.
>
> But I would prefer:
> #define SET 4
> #define CLEAR 8
> [...]
> writel(CTRL_DOTCLK_MODE, host->base + CTRL + SET)
> :-)
>
> Regards,
> Juergen
^ permalink raw reply [flat|nested] 65+ messages in thread
* [PATCH 1/2] video: Add i.MX23/28 framebuffer driver
@ 2011-02-10 10:08 ` Li Frank-B20596
0 siblings, 0 replies; 65+ messages in thread
From: Li Frank-B20596 @ 2011-02-10 10:08 UTC (permalink / raw)
To: linux-arm-kernel
> Li Frank-B20596 wrote:
> > > +
> > > + /* if it was disabled, re-enable the mode again */
> > > + reg = readl(host->base + CTRL);
> > > + reg |= CTRL_DOTCLK_MODE;
> > > + writel(reg, host->base + CTRL);
> >
> > writel(CTRL_DOTCLK_MODE, host->base + CTRL_SET)
> > CTRL_SET is CTRL+0x4
> > SET and CLR register is easier than Read and write back.
>
> But you must always check, if the register really has such a SET and
> CLEAR
> feature. Not all registers have this feature...
That's a reason why I suggest use register header that generate from xml.
HW_XX_CTRL_SET will exist if there are such register in hardware.
If hardware have not such register, there are not _SET and _CLR.
Compiler will report error if user use _SET but hardware not such register.
>
> But in this case you are right.
>
> But I would prefer:
> #define SET 4
> #define CLEAR 8
> [...]
> writel(CTRL_DOTCLK_MODE, host->base + CTRL + SET)
> :-)
>
> Regards,
> Juergen
^ permalink raw reply [flat|nested] 65+ messages in thread
* RE: [PATCH 1/2] video: Add i.MX23/28 framebuffer driver
2011-02-10 9:52 ` Juergen Beisert
@ 2011-02-10 10:37 ` Li Frank-B20596
-1 siblings, 0 replies; 65+ messages in thread
From: Li Frank-B20596 @ 2011-02-10 10:37 UTC (permalink / raw)
To: linux-arm-kernel
> This kind of macro encryption _may_ help when you are coding the driver
> the
> first time. But after reading and reading it again (while testing and
> debugging) all these prefixes and suffixes do not add any valuable
> information. They only fill up your lines, enlarges your source code and
> make
> you blind for the real bug...
>
> Everyone who wants to see how source code looks that uses these longs
> macros I
> can recommend reading the so called 'bootlets' source code. :-))
The complex of bootlets is not long macro, but the work around to make sure
Chip PMU work safely. If there are not such macro, power_prep will become
more difficult to understand.
>
> > Developer needn't look up register header file when coding, just
>
> That's why I add these macros into the source file instead into a header
> file.
>
> > write down Register name or bit name according to mx23/mx28 spec.
>
> And sometimes the spec is incomplete or just wrong and so on. Independent
> of
> any macro naming policy.
Not at all. Mx23/mx28 can keep spec, header file and RTL consistent just
because using one source and multiple output. During Freescale mx28 BSP
develop, never happen bit and register definition error.
>
> > If you still think it is too long, I suggest keep HW_ and BM_ prefix to
> > distinguish Which one is register name, which one is bit mask.
>
> The used macro in the address part of the writel() must be an offset,
> while
> the macro used in the value part must be a bit definition. Anything else
> is
> redundant.
If name is too short, there are possible to cause name pollution.
BM_, BP_, BF_ have their implication.
VDCTRL4_SET_DOTCLK_DLY(x), according to our macro policy, it should be
BF_VDCTRL4_SET_DOTCLK_DLY(x).
Generally, you need
Reg = readl(offset+VDCTRL4);
Reg &=~ ((0x7)<<29);
Reg |= VDCTRL4_SET_DOTCLK_DLY(value)
Writel(reg, offset+VDCTRL4)
Compared with our macro policy.
Reg = readl(offset+HW_VDCTRL4);
Reg &=~ BM_VDCTRL4_SET_DOTCLK_DLY;
Reg |= BF_VDCTRL4_SET_DOTCLK_DLY(value)
Writel(reg, offset+HW_VDCTRL4)
>
> Regards,
> Juergen
>
^ permalink raw reply [flat|nested] 65+ messages in thread
* [PATCH 1/2] video: Add i.MX23/28 framebuffer driver
@ 2011-02-10 10:37 ` Li Frank-B20596
0 siblings, 0 replies; 65+ messages in thread
From: Li Frank-B20596 @ 2011-02-10 10:37 UTC (permalink / raw)
To: linux-arm-kernel
> This kind of macro encryption _may_ help when you are coding the driver
> the
> first time. But after reading and reading it again (while testing and
> debugging) all these prefixes and suffixes do not add any valuable
> information. They only fill up your lines, enlarges your source code and
> make
> you blind for the real bug...
>
> Everyone who wants to see how source code looks that uses these longs
> macros I
> can recommend reading the so called 'bootlets' source code. :-))
The complex of bootlets is not long macro, but the work around to make sure
Chip PMU work safely. If there are not such macro, power_prep will become
more difficult to understand.
>
> > Developer needn't look up register header file when coding, just
>
> That's why I add these macros into the source file instead into a header
> file.
>
> > write down Register name or bit name according to mx23/mx28 spec.
>
> And sometimes the spec is incomplete or just wrong and so on. Independent
> of
> any macro naming policy.
Not at all. Mx23/mx28 can keep spec, header file and RTL consistent just
because using one source and multiple output. During Freescale mx28 BSP
develop, never happen bit and register definition error.
>
> > If you still think it is too long, I suggest keep HW_ and BM_ prefix to
> > distinguish Which one is register name, which one is bit mask.
>
> The used macro in the address part of the writel() must be an offset,
> while
> the macro used in the value part must be a bit definition. Anything else
> is
> redundant.
If name is too short, there are possible to cause name pollution.
BM_, BP_, BF_ have their implication.
VDCTRL4_SET_DOTCLK_DLY(x), according to our macro policy, it should be
BF_VDCTRL4_SET_DOTCLK_DLY(x).
Generally, you need
Reg = readl(offset+VDCTRL4);
Reg &=~ ((0x7)<<29);
Reg |= VDCTRL4_SET_DOTCLK_DLY(value)
Writel(reg, offset+VDCTRL4)
Compared with our macro policy.
Reg = readl(offset+HW_VDCTRL4);
Reg &=~ BM_VDCTRL4_SET_DOTCLK_DLY;
Reg |= BF_VDCTRL4_SET_DOTCLK_DLY(value)
Writel(reg, offset+HW_VDCTRL4)
>
> Regards,
> Juergen
>
^ permalink raw reply [flat|nested] 65+ messages in thread
* Re: [PATCH 1/2] video: Add i.MX23/28 framebuffer driver
2011-02-10 9:46 ` Juergen Beisert
@ 2011-02-10 10:47 ` Wolfram Sang
-1 siblings, 0 replies; 65+ messages in thread
From: Wolfram Sang @ 2011-02-10 10:47 UTC (permalink / raw)
To: linux-arm-kernel
[-- Attachment #1: Type: text/plain, Size: 517 bytes --]
> But I would prefer:
> #define SET 4
> #define CLEAR 8
> [...]
> writel(CTRL_DOTCLK_MODE, host->base + CTRL + SET)
NACK.
This will indeed not be caught by the compiler.
And while I like the shorter macros better as well, an MXSFB_ prefix
should be added IMO. Macros like CTRL and TIMING are not too well
protected.
Regards,
Wolfram
--
Pengutronix e.K. | Wolfram Sang |
Industrial Linux Solutions | http://www.pengutronix.de/ |
[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 198 bytes --]
^ permalink raw reply [flat|nested] 65+ messages in thread
* [PATCH 1/2] video: Add i.MX23/28 framebuffer driver
@ 2011-02-10 10:47 ` Wolfram Sang
0 siblings, 0 replies; 65+ messages in thread
From: Wolfram Sang @ 2011-02-10 10:47 UTC (permalink / raw)
To: linux-arm-kernel
> But I would prefer:
> #define SET 4
> #define CLEAR 8
> [...]
> writel(CTRL_DOTCLK_MODE, host->base + CTRL + SET)
NACK.
This will indeed not be caught by the compiler.
And while I like the shorter macros better as well, an MXSFB_ prefix
should be added IMO. Macros like CTRL and TIMING are not too well
protected.
Regards,
Wolfram
--
Pengutronix e.K. | Wolfram Sang |
Industrial Linux Solutions | http://www.pengutronix.de/ |
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 198 bytes
Desc: Digital signature
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20110210/d36ddd33/attachment.sig>
^ permalink raw reply [flat|nested] 65+ messages in thread
* Re: [PATCH 1/2] video: Add i.MX23/28 framebuffer driver
2011-02-10 10:37 ` Li Frank-B20596
@ 2011-02-10 11:12 ` Juergen Beisert
-1 siblings, 0 replies; 65+ messages in thread
From: Juergen Beisert @ 2011-02-10 11:12 UTC (permalink / raw)
To: linux-arm-kernel
Li Frank-B20596 wrote:
> > This kind of macro encryption _may_ help when you are coding the driver
> > the
> > first time. But after reading and reading it again (while testing and
> > debugging) all these prefixes and suffixes do not add any valuable
> > information. They only fill up your lines, enlarges your source code and
> > make
> > you blind for the real bug...
> >
> > Everyone who wants to see how source code looks that uses these longs
> > macros I
> > can recommend reading the so called 'bootlets' source code. :-))
>
> The complex of bootlets is not long macro, but the work around to make sure
> Chip PMU work safely. If there are not such macro, power_prep will become
> more difficult to understand.
Its your opinion...
> > > Developer needn't look up register header file when coding, just
> >
> > That's why I add these macros into the source file instead into a header
> > file.
> >
> > > write down Register name or bit name according to mx23/mx28 spec.
> >
> > And sometimes the spec is incomplete or just wrong and so on. Independent
> > of any macro naming policy.
>
> Not at all. Mx23/mx28 can keep spec, header file and RTL consistent just
> because using one source and multiple output. During Freescale mx28 BSP
> develop, never happen bit and register definition error.
>
> > > If you still think it is too long, I suggest keep HW_ and BM_ prefix to
> > > distinguish Which one is register name, which one is bit mask.
> >
> > The used macro in the address part of the writel() must be an offset,
> > while
> > the macro used in the value part must be a bit definition. Anything else
> > is
> > redundant.
>
> If name is too short, there are possible to cause name pollution.
> BM_, BP_, BF_ have their implication.
If you collect everything in header files: Yes. If you keep the macros where
they are used, IMHO: No.
> VDCTRL4_SET_DOTCLK_DLY(x), according to our macro policy, it should be
> BF_VDCTRL4_SET_DOTCLK_DLY(x).
>
> Generally, you need
> Reg = readl(offset+VDCTRL4);
> Reg &=~ ((0x7)<<29);
What about?
> Reg &= ~VDCTRL4_SET_DOTCLK_DLY(0x7);
Or?
> Reg &= ~VDCTRL4_SET_DOTCLK_DLY_MASK;
Okay, the even get longer and longer.
> Reg |= VDCTRL4_SET_DOTCLK_DLY(value)
> Writel(reg, offset+VDCTRL4)
>
> Compared with our macro policy.
>
> Reg = readl(offset+HW_VDCTRL4);
> Reg &=~ BM_VDCTRL4_SET_DOTCLK_DLY;
> Reg |= BF_VDCTRL4_SET_DOTCLK_DLY(value)
> Writel(reg, offset+HW_VDCTRL4)
Anywhere else such naming policy in the kernel? I have to deal with all kind
of CPUs, not only i.MX23/28.
Regards,
Juergen
--
Pengutronix e.K. | Juergen Beisert |
Linux Solutions for Science and Industry | Phone: +49-8766-939 228 |
Vertretung Sued/Muenchen, Germany | Fax: +49-5121-206917-5555 |
Amtsgericht Hildesheim, HRA 2686 | http://www.pengutronix.de/ |
^ permalink raw reply [flat|nested] 65+ messages in thread
* [PATCH 1/2] video: Add i.MX23/28 framebuffer driver
@ 2011-02-10 11:12 ` Juergen Beisert
0 siblings, 0 replies; 65+ messages in thread
From: Juergen Beisert @ 2011-02-10 11:12 UTC (permalink / raw)
To: linux-arm-kernel
Li Frank-B20596 wrote:
> > This kind of macro encryption _may_ help when you are coding the driver
> > the
> > first time. But after reading and reading it again (while testing and
> > debugging) all these prefixes and suffixes do not add any valuable
> > information. They only fill up your lines, enlarges your source code and
> > make
> > you blind for the real bug...
> >
> > Everyone who wants to see how source code looks that uses these longs
> > macros I
> > can recommend reading the so called 'bootlets' source code. :-))
>
> The complex of bootlets is not long macro, but the work around to make sure
> Chip PMU work safely. If there are not such macro, power_prep will become
> more difficult to understand.
Its your opinion...
> > > Developer needn't look up register header file when coding, just
> >
> > That's why I add these macros into the source file instead into a header
> > file.
> >
> > > write down Register name or bit name according to mx23/mx28 spec.
> >
> > And sometimes the spec is incomplete or just wrong and so on. Independent
> > of any macro naming policy.
>
> Not at all. Mx23/mx28 can keep spec, header file and RTL consistent just
> because using one source and multiple output. During Freescale mx28 BSP
> develop, never happen bit and register definition error.
>
> > > If you still think it is too long, I suggest keep HW_ and BM_ prefix to
> > > distinguish Which one is register name, which one is bit mask.
> >
> > The used macro in the address part of the writel() must be an offset,
> > while
> > the macro used in the value part must be a bit definition. Anything else
> > is
> > redundant.
>
> If name is too short, there are possible to cause name pollution.
> BM_, BP_, BF_ have their implication.
If you collect everything in header files: Yes. If you keep the macros where
they are used, IMHO: No.
> VDCTRL4_SET_DOTCLK_DLY(x), according to our macro policy, it should be
> BF_VDCTRL4_SET_DOTCLK_DLY(x).
>
> Generally, you need
> Reg = readl(offset+VDCTRL4);
> Reg &=~ ((0x7)<<29);
What about?
> Reg &= ~VDCTRL4_SET_DOTCLK_DLY(0x7);
Or?
> Reg &= ~VDCTRL4_SET_DOTCLK_DLY_MASK;
Okay, the even get longer and longer.
> Reg |= VDCTRL4_SET_DOTCLK_DLY(value)
> Writel(reg, offset+VDCTRL4)
>
> Compared with our macro policy.
>
> Reg = readl(offset+HW_VDCTRL4);
> Reg &=~ BM_VDCTRL4_SET_DOTCLK_DLY;
> Reg |= BF_VDCTRL4_SET_DOTCLK_DLY(value)
> Writel(reg, offset+HW_VDCTRL4)
Anywhere else such naming policy in the kernel? I have to deal with all kind
of CPUs, not only i.MX23/28.
Regards,
Juergen
--
Pengutronix e.K. | Juergen Beisert |
Linux Solutions for Science and Industry | Phone: +49-8766-939 228 |
Vertretung Sued/Muenchen, Germany | Fax: +49-5121-206917-5555 |
Amtsgericht Hildesheim, HRA 2686 | http://www.pengutronix.de/ |
^ permalink raw reply [flat|nested] 65+ messages in thread
* Re: [PATCH 1/2] video: Add i.MX23/28 framebuffer driver
2011-02-10 10:47 ` Wolfram Sang
@ 2011-02-10 11:51 ` Sascha Hauer
-1 siblings, 0 replies; 65+ messages in thread
From: Sascha Hauer @ 2011-02-10 11:51 UTC (permalink / raw)
To: linux-arm-kernel
On Thu, Feb 10, 2011 at 11:47:25AM +0100, Wolfram Sang wrote:
> > But I would prefer:
> > #define SET 4
> > #define CLEAR 8
> > [...]
> > writel(CTRL_DOTCLK_MODE, host->base + CTRL + SET)
For this we already have __mxs_setl.
>
> NACK.
>
> This will indeed not be caught by the compiler.
>
> And while I like the shorter macros better as well, an MXSFB_ prefix
> should be added IMO. Macros like CTRL and TIMING are not too well
> protected.
+1 for a prefix, be it MXSFB_ or LCDIF_, will change this.
Otherwise can we please stop about arguing about register defines? It's
too much a matter of taste and there is no perfect way to solve this.
Sascha
--
Pengutronix e.K. | |
Industrial Linux Solutions | http://www.pengutronix.de/ |
Peiner Str. 6-8, 31137 Hildesheim, Germany | Phone: +49-5121-206917-0 |
Amtsgericht Hildesheim, HRA 2686 | Fax: +49-5121-206917-5555 |
^ permalink raw reply [flat|nested] 65+ messages in thread
* [PATCH 1/2] video: Add i.MX23/28 framebuffer driver
@ 2011-02-10 11:51 ` Sascha Hauer
0 siblings, 0 replies; 65+ messages in thread
From: Sascha Hauer @ 2011-02-10 11:51 UTC (permalink / raw)
To: linux-arm-kernel
On Thu, Feb 10, 2011 at 11:47:25AM +0100, Wolfram Sang wrote:
> > But I would prefer:
> > #define SET 4
> > #define CLEAR 8
> > [...]
> > writel(CTRL_DOTCLK_MODE, host->base + CTRL + SET)
For this we already have __mxs_setl.
>
> NACK.
>
> This will indeed not be caught by the compiler.
>
> And while I like the shorter macros better as well, an MXSFB_ prefix
> should be added IMO. Macros like CTRL and TIMING are not too well
> protected.
+1 for a prefix, be it MXSFB_ or LCDIF_, will change this.
Otherwise can we please stop about arguing about register defines? It's
too much a matter of taste and there is no perfect way to solve this.
Sascha
--
Pengutronix e.K. | |
Industrial Linux Solutions | http://www.pengutronix.de/ |
Peiner Str. 6-8, 31137 Hildesheim, Germany | Phone: +49-5121-206917-0 |
Amtsgericht Hildesheim, HRA 2686 | Fax: +49-5121-206917-5555 |
^ permalink raw reply [flat|nested] 65+ messages in thread
* RE: [PATCH 1/2] video: Add i.MX23/28 framebuffer driver
2011-02-10 11:12 ` Juergen Beisert
@ 2011-02-10 12:23 ` Li Frank-B20596
-1 siblings, 0 replies; 65+ messages in thread
From: Li Frank-B20596 @ 2011-02-10 12:23 UTC (permalink / raw)
To: linux-arm-kernel
>
> What about?
> > Reg &= ~VDCTRL4_SET_DOTCLK_DLY(0x7);
> Or?
> > Reg &= ~VDCTRL4_SET_DOTCLK_DLY_MASK;
> Okay, the even get longer and longer.
If use 0x7, you have to take care bit width.
If use VDCTRL4_SET_DOTCLK_DLY_MASK, I think BF_ and BM_ is better.
> > Reg |= VDCTRL4_SET_DOTCLK_DLY(value)
> > Writel(reg, offset+VDCTRL4)
> >
> > Compared with our macro policy.
> >
> > Reg = readl(offset+HW_VDCTRL4);
> > Reg &=~ BM_VDCTRL4_SET_DOTCLK_DLY;
> > Reg |= BF_VDCTRL4_SET_DOTCLK_DLY(value)
> > Writel(reg, offset+HW_VDCTRL4)
>
> Anywhere else such naming policy in the kernel? I have to deal with all
> kind
> of CPUs, not only i.MX23/28.
This driver is for mx23/mx28.
I have not found such policy in the kernel, otherwise I have generated that at beginning.
It is better if all of mx23/mx28 driver follow the same role.
^ permalink raw reply [flat|nested] 65+ messages in thread
* [PATCH 1/2] video: Add i.MX23/28 framebuffer driver
@ 2011-02-10 12:23 ` Li Frank-B20596
0 siblings, 0 replies; 65+ messages in thread
From: Li Frank-B20596 @ 2011-02-10 12:23 UTC (permalink / raw)
To: linux-arm-kernel
>
> What about?
> > Reg &= ~VDCTRL4_SET_DOTCLK_DLY(0x7);
> Or?
> > Reg &= ~VDCTRL4_SET_DOTCLK_DLY_MASK;
> Okay, the even get longer and longer.
If use 0x7, you have to take care bit width.
If use VDCTRL4_SET_DOTCLK_DLY_MASK, I think BF_ and BM_ is better.
> > Reg |= VDCTRL4_SET_DOTCLK_DLY(value)
> > Writel(reg, offset+VDCTRL4)
> >
> > Compared with our macro policy.
> >
> > Reg = readl(offset+HW_VDCTRL4);
> > Reg &=~ BM_VDCTRL4_SET_DOTCLK_DLY;
> > Reg |= BF_VDCTRL4_SET_DOTCLK_DLY(value)
> > Writel(reg, offset+HW_VDCTRL4)
>
> Anywhere else such naming policy in the kernel? I have to deal with all
> kind
> of CPUs, not only i.MX23/28.
This driver is for mx23/mx28.
I have not found such policy in the kernel, otherwise I have generated that at beginning.
It is better if all of mx23/mx28 driver follow the same role.
^ permalink raw reply [flat|nested] 65+ messages in thread
* RE: [PATCH 1/2] video: Add i.MX23/28 framebuffer driver
2011-02-10 11:51 ` Sascha Hauer
@ 2011-02-10 12:32 ` Li Frank-B20596
-1 siblings, 0 replies; 65+ messages in thread
From: Li Frank-B20596 @ 2011-02-10 12:32 UTC (permalink / raw)
To: linux-arm-kernel
> On Thu, Feb 10, 2011 at 11:47:25AM +0100, Wolfram Sang wrote:
> > > But I would prefer:
> > > #define SET 4
> > > #define CLEAR 8
> > > [...]
> > > writel(CTRL_DOTCLK_MODE, host->base + CTRL + SET)
>
> For this we already have __mxs_setl.
>
> >
> > NACK.
> >
> > This will indeed not be caught by the compiler.
> >
> > And while I like the shorter macros better as well, an MXSFB_ prefix
> > should be added IMO. Macros like CTRL and TIMING are not too well
> > protected.
>
> +1 for a prefix, be it MXSFB_ or LCDIF_, will change this.
>
> Otherwise can we please stop about arguing about register defines? It's
> too much a matter of taste and there is no perfect way to solve this.
>
> Sascha
>
Guo Shawn mmc driver's patch is good example to use mx23/mx28 register macro.
Best regards
Frank Li
^ permalink raw reply [flat|nested] 65+ messages in thread
* [PATCH 1/2] video: Add i.MX23/28 framebuffer driver
@ 2011-02-10 12:32 ` Li Frank-B20596
0 siblings, 0 replies; 65+ messages in thread
From: Li Frank-B20596 @ 2011-02-10 12:32 UTC (permalink / raw)
To: linux-arm-kernel
> On Thu, Feb 10, 2011 at 11:47:25AM +0100, Wolfram Sang wrote:
> > > But I would prefer:
> > > #define SET 4
> > > #define CLEAR 8
> > > [...]
> > > writel(CTRL_DOTCLK_MODE, host->base + CTRL + SET)
>
> For this we already have __mxs_setl.
>
> >
> > NACK.
> >
> > This will indeed not be caught by the compiler.
> >
> > And while I like the shorter macros better as well, an MXSFB_ prefix
> > should be added IMO. Macros like CTRL and TIMING are not too well
> > protected.
>
> +1 for a prefix, be it MXSFB_ or LCDIF_, will change this.
>
> Otherwise can we please stop about arguing about register defines? It's
> too much a matter of taste and there is no perfect way to solve this.
>
> Sascha
>
Guo Shawn mmc driver's patch is good example to use mx23/mx28 register macro.
Best regards
Frank Li
^ permalink raw reply [flat|nested] 65+ messages in thread
* Re: [PATCH] i.MX23/28 framebuffer driver
2011-02-09 16:31 ` Arnd Bergmann
@ 2011-02-10 17:09 ` Robert Schwebel
-1 siblings, 0 replies; 65+ messages in thread
From: Robert Schwebel @ 2011-02-10 17:09 UTC (permalink / raw)
To: linux-arm-kernel
On Wed, Feb 09, 2011 at 05:31:07PM +0100, Arnd Bergmann wrote:
> Ok. This sounds like a lot of upfront work indeed, to make KMS more
> generic, though I think a number of driver would benefit from it
> eventually. It could be something for the Linaro graphics working
> group to look at in the following 11.11 release, depending on how
> many other people are interested in getting there.
There is already a blueprint at linaro:
https://blueprints.launchpad.net/linaro-graphics-wg/+spec/multimedia-linaro-dri-1105
We have recently tested the state of Qt acceleration on the MX51, in
order to find out which is the best way to make use of the
possibilities. However, at least at the moment this requires highly
experimental Qt sceenegraph git branches, and even then it results in
lower perfomance than with software rendering ...
But for the long term, the linaro blueprint looks like the way to go.
rsc
--
Pengutronix e.K. | |
Industrial Linux Solutions | http://www.pengutronix.de/ |
Peiner Str. 6-8, 31137 Hildesheim, Germany | Phone: +49-5121-206917-0 |
Amtsgericht Hildesheim, HRA 2686 | Fax: +49-5121-206917-5555 |
^ permalink raw reply [flat|nested] 65+ messages in thread
* [PATCH] i.MX23/28 framebuffer driver
@ 2011-02-10 17:09 ` Robert Schwebel
0 siblings, 0 replies; 65+ messages in thread
From: Robert Schwebel @ 2011-02-10 17:09 UTC (permalink / raw)
To: linux-arm-kernel
On Wed, Feb 09, 2011 at 05:31:07PM +0100, Arnd Bergmann wrote:
> Ok. This sounds like a lot of upfront work indeed, to make KMS more
> generic, though I think a number of driver would benefit from it
> eventually. It could be something for the Linaro graphics working
> group to look at in the following 11.11 release, depending on how
> many other people are interested in getting there.
There is already a blueprint at linaro:
https://blueprints.launchpad.net/linaro-graphics-wg/+spec/multimedia-linaro-dri-1105
We have recently tested the state of Qt acceleration on the MX51, in
order to find out which is the best way to make use of the
possibilities. However, at least at the moment this requires highly
experimental Qt sceenegraph git branches, and even then it results in
lower perfomance than with software rendering ...
But for the long term, the linaro blueprint looks like the way to go.
rsc
--
Pengutronix e.K. | |
Industrial Linux Solutions | http://www.pengutronix.de/ |
Peiner Str. 6-8, 31137 Hildesheim, Germany | Phone: +49-5121-206917-0 |
Amtsgericht Hildesheim, HRA 2686 | Fax: +49-5121-206917-5555 |
^ permalink raw reply [flat|nested] 65+ messages in thread
* Re: [PATCH] i.MX23/28 framebuffer driver
2011-02-09 16:31 ` Arnd Bergmann
@ 2011-02-15 14:13 ` Clark, Rob
-1 siblings, 0 replies; 65+ messages in thread
From: Clark, Rob @ 2011-02-15 14:13 UTC (permalink / raw)
To: linux-arm-kernel
On Wed, Feb 9, 2011 at 10:31 AM, Arnd Bergmann <arnd@arndb.de> wrote:
> On Wednesday 09 February 2011, Sascha Hauer wrote:
>
>> The driver patch itself is Cced to linux-fbdev, only the introductory
>> mail is not.
>
> Ok, I see.
>
>> > Did you consider making the driver a KMS driver instead of
>> > a frame buffer? I think the recommendation these days is
>> > to start out with KMS for new drivers, which will be somewhat
>> > simpler and give you a frame buffer device as well. I don't
>> > think that there is a need to change over any drivers from
>> > fb to kms though, since you've already done the work.
>>
>> I tried doing so for the i.MX51 which supports multiple displays on dvi
>> and vga outputs and thus could make good use of KMS and friends. Anyway,
>> I got stuck quite fast. The KMS stuff is tightly coupled with DRM/DRI
>> and needs many many callbacks to implement. Additionally the userspace
>> tools expect a nvidia/amd/intel driver and do not have a generic
>> fallback. I think this stuff is good for implementing a full blown
>> graphics driver, but is lacking support for simple framebuffer grapics.
>> I'd love to go this way but it still requires a lot of work.
>
> Ok. This sounds like a lot of upfront work indeed, to make KMS more
> generic, though I think a number of driver would benefit from it
> eventually. It could be something for the Linaro graphics working
> group to look at in the following 11.11 release, depending on how
> many other people are interested in getting there.
>
fwiw, it seems to me like xorg could have some more common code to
deal with the KMS part of DRM.. the various userspace xorg drivers end
up having a lot of very similar code to deal with enumerating
CRTCs/outputs and modes, handle hotplug, etc.
I'd been experimenting a bit on the side w/ the DRM driver framework (
http://gitorious.com/~robclark/pandaboard/robclarks-kernel-omap4/commits/omap_gpu
), but had to add a good chunk of mostly boilerplate code to our xorg
driver in order just to test it. Maybe some generic support for KMS
in xf86-video-fbdev would have made this easier to develop the kernel
part without in parallel having to implement the userspace part. I'm
not sure if this is the sort of thing the linaro-wg has in mind?
BR,
-R
^ permalink raw reply [flat|nested] 65+ messages in thread
* [PATCH] i.MX23/28 framebuffer driver
@ 2011-02-15 14:13 ` Clark, Rob
0 siblings, 0 replies; 65+ messages in thread
From: Clark, Rob @ 2011-02-15 14:13 UTC (permalink / raw)
To: linux-arm-kernel
On Wed, Feb 9, 2011 at 10:31 AM, Arnd Bergmann <arnd@arndb.de> wrote:
> On Wednesday 09 February 2011, Sascha Hauer wrote:
>
>> The driver patch itself is Cced to linux-fbdev, only the introductory
>> mail is not.
>
> Ok, I see.
>
>> > Did you consider making the driver a KMS driver instead of
>> > a frame buffer? I think the recommendation these days is
>> > to start out with KMS for new drivers, which will be somewhat
>> > simpler and give you a frame buffer device as well. I don't
>> > think that there is a need to change over any drivers from
>> > fb to kms though, since you've already done the work.
>>
>> I tried doing so for the i.MX51 which supports multiple displays on dvi
>> and vga outputs and thus could make good use of KMS and friends. Anyway,
>> I got stuck quite fast. The KMS stuff is tightly coupled with DRM/DRI
>> and needs many many callbacks to implement. Additionally the userspace
>> tools expect a nvidia/amd/intel driver and do not have a generic
>> fallback. I think this stuff is good for implementing a full blown
>> graphics driver, but is lacking support for simple framebuffer grapics.
>> I'd love to go this way but it still requires a lot of work.
>
> Ok. This sounds like a lot of upfront work indeed, to make KMS more
> generic, though I think a number of driver would benefit from it
> eventually. It could be something for the Linaro graphics working
> group to look at in the following 11.11 release, depending on how
> many other people are interested in getting there.
>
fwiw, it seems to me like xorg could have some more common code to
deal with the KMS part of DRM.. the various userspace xorg drivers end
up having a lot of very similar code to deal with enumerating
CRTCs/outputs and modes, handle hotplug, etc.
I'd been experimenting a bit on the side w/ the DRM driver framework (
http://gitorious.com/~robclark/pandaboard/robclarks-kernel-omap4/commits/omap_gpu
), but had to add a good chunk of mostly boilerplate code to our xorg
driver in order just to test it. Maybe some generic support for KMS
in xf86-video-fbdev would have made this easier to develop the kernel
part without in parallel having to implement the userspace part. I'm
not sure if this is the sort of thing the linaro-wg has in mind?
BR,
-R
^ permalink raw reply [flat|nested] 65+ messages in thread
* Re: [PATCH] i.MX23/28 framebuffer driver
2011-02-15 14:13 ` Clark, Rob
@ 2011-02-16 12:22 ` Arnd Bergmann
-1 siblings, 0 replies; 65+ messages in thread
From: Arnd Bergmann @ 2011-02-16 12:22 UTC (permalink / raw)
To: linux-arm-kernel
On Tuesday 15 February 2011, Clark, Rob wrote:
> I'd been experimenting a bit on the side w/ the DRM driver framework (
> http://gitorious.com/~robclark/pandaboard/robclarks-kernel-omap4/commits/omap_gpu
> ), but had to add a good chunk of mostly boilerplate code to our xorg
> driver in order just to test it. Maybe some generic support for KMS
> in xf86-video-fbdev would have made this easier to develop the kernel
> part without in parallel having to implement the userspace part. I'm
> not sure if this is the sort of thing the linaro-wg has in mind?
I'm not sure what the the linaro multimedia wg thinks of this, but the
kernel code you linked looks like it's doing exactly the right thing.
Arnd
^ permalink raw reply [flat|nested] 65+ messages in thread
* [PATCH] i.MX23/28 framebuffer driver
@ 2011-02-16 12:22 ` Arnd Bergmann
0 siblings, 0 replies; 65+ messages in thread
From: Arnd Bergmann @ 2011-02-16 12:22 UTC (permalink / raw)
To: linux-arm-kernel
On Tuesday 15 February 2011, Clark, Rob wrote:
> I'd been experimenting a bit on the side w/ the DRM driver framework (
> http://gitorious.com/~robclark/pandaboard/robclarks-kernel-omap4/commits/omap_gpu
> ), but had to add a good chunk of mostly boilerplate code to our xorg
> driver in order just to test it. Maybe some generic support for KMS
> in xf86-video-fbdev would have made this easier to develop the kernel
> part without in parallel having to implement the userspace part. I'm
> not sure if this is the sort of thing the linaro-wg has in mind?
I'm not sure what the the linaro multimedia wg thinks of this, but the
kernel code you linked looks like it's doing exactly the right thing.
Arnd
^ permalink raw reply [flat|nested] 65+ messages in thread
* [PATCH] i.MX23/28 framebuffer driver
2011-02-16 12:22 ` Arnd Bergmann
(?)
@ 2011-02-16 14:42 ` Jesse Barker
2011-02-17 1:08 ` Clark, Rob
-1 siblings, 1 reply; 65+ messages in thread
From: Jesse Barker @ 2011-02-16 14:42 UTC (permalink / raw)
To: linux-arm-kernel
Speaking for the Linaro graphics working group, I think it's great. And, I
think you're right, that if enough of the KMS support in xf86-video-* is
similar enough (I was only aware of intel and nouveau supporting it properly
at current), pulling it out into a common layer would make it easier to
support in new drivers (including fbdev).
cheers,
Jesse
On Wed, Feb 16, 2011 at 4:22 AM, Arnd Bergmann <arnd@arndb.de> wrote:
> On Tuesday 15 February 2011, Clark, Rob wrote:
> > I'd been experimenting a bit on the side w/ the DRM driver framework (
> >
> http://gitorious.com/~robclark/pandaboard/robclarks-kernel-omap4/commits/omap_gpu
> > ), but had to add a good chunk of mostly boilerplate code to our xorg
> > driver in order just to test it. Maybe some generic support for KMS
> > in xf86-video-fbdev would have made this easier to develop the kernel
> > part without in parallel having to implement the userspace part. I'm
> > not sure if this is the sort of thing the linaro-wg has in mind?
>
> I'm not sure what the the linaro multimedia wg thinks of this, but the
> kernel code you linked looks like it's doing exactly the right thing.
>
> Arnd
>
> _______________________________________________
> linaro-dev mailing list
> linaro-dev at lists.linaro.org
> http://lists.linaro.org/mailman/listinfo/linaro-dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20110216/78490466/attachment-0001.html>
^ permalink raw reply [flat|nested] 65+ messages in thread
* Re: [PATCH] i.MX23/28 framebuffer driver
2011-02-16 14:42 ` Jesse Barker
@ 2011-02-17 1:08 ` Clark, Rob
0 siblings, 0 replies; 65+ messages in thread
From: Clark, Rob @ 2011-02-17 1:08 UTC (permalink / raw)
To: linux-arm-kernel
I'm still in the learning-as-I-go phase here, so definitely not ready
to propose a solution, but it does seem to me like there is room for
some sort of kms-helper library here to handle more of the boilerplate
xf86-video-* stuff.. I guess I'll have a better picture once I have a
chance to add support for the various multi-monitor configurations.
But certainly would be interested if anyone already has some ideas.
BR,
-R
On Wed, Feb 16, 2011 at 8:42 AM, Jesse Barker <jesse.barker@linaro.org> wrote:
> Speaking for the Linaro graphics working group, I think it's great. And, I
> think you're right, that if enough of the KMS support in xf86-video-* is
> similar enough (I was only aware of intel and nouveau supporting it properly
> at current), pulling it out into a common layer would make it easier to
> support in new drivers (including fbdev).
>
> cheers,
> Jesse
>
> On Wed, Feb 16, 2011 at 4:22 AM, Arnd Bergmann <arnd@arndb.de> wrote:
>>
>> On Tuesday 15 February 2011, Clark, Rob wrote:
>> > I'd been experimenting a bit on the side w/ the DRM driver framework (
>> >
>> > http://gitorious.com/~robclark/pandaboard/robclarks-kernel-omap4/commits/omap_gpu
>> > ), but had to add a good chunk of mostly boilerplate code to our xorg
>> > driver in order just to test it. Maybe some generic support for KMS
>> > in xf86-video-fbdev would have made this easier to develop the kernel
>> > part without in parallel having to implement the userspace part. I'm
>> > not sure if this is the sort of thing the linaro-wg has in mind?
>>
>> I'm not sure what the the linaro multimedia wg thinks of this, but the
>> kernel code you linked looks like it's doing exactly the right thing.
>>
>> Arnd
>>
>> _______________________________________________
>> linaro-dev mailing list
>> linaro-dev@lists.linaro.org
>> http://lists.linaro.org/mailman/listinfo/linaro-dev
>
>
^ permalink raw reply [flat|nested] 65+ messages in thread
* [PATCH] i.MX23/28 framebuffer driver
@ 2011-02-17 1:08 ` Clark, Rob
0 siblings, 0 replies; 65+ messages in thread
From: Clark, Rob @ 2011-02-17 1:08 UTC (permalink / raw)
To: linux-arm-kernel
I'm still in the learning-as-I-go phase here, so definitely not ready
to propose a solution, but it does seem to me like there is room for
some sort of kms-helper library here to handle more of the boilerplate
xf86-video-* stuff.. I guess I'll have a better picture once I have a
chance to add support for the various multi-monitor configurations.
But certainly would be interested if anyone already has some ideas.
BR,
-R
On Wed, Feb 16, 2011 at 8:42 AM, Jesse Barker <jesse.barker@linaro.org> wrote:
> Speaking for the Linaro graphics working group, I think it's great.? And, I
> think you're right, that if enough of the KMS support in xf86-video-* is
> similar enough (I was only aware of intel and nouveau supporting it properly
> at current), pulling it out into a common layer would make it easier to
> support in new drivers (including fbdev).
>
> cheers,
> Jesse
>
> On Wed, Feb 16, 2011 at 4:22 AM, Arnd Bergmann <arnd@arndb.de> wrote:
>>
>> On Tuesday 15 February 2011, Clark, Rob wrote:
>> > I'd been experimenting a bit on the side w/ the DRM driver framework (
>> >
>> > http://gitorious.com/~robclark/pandaboard/robclarks-kernel-omap4/commits/omap_gpu
>> > ), but had to add a good chunk of mostly boilerplate code to our xorg
>> > driver in order just to test it. ?Maybe some generic support for KMS
>> > in xf86-video-fbdev would have made this easier to develop the kernel
>> > part without in parallel having to implement the userspace part. ?I'm
>> > not sure if this is the sort of thing the linaro-wg has in mind?
>>
>> I'm not sure what the the linaro multimedia wg thinks of this, but the
>> kernel code you linked looks like it's doing exactly the right thing.
>>
>> ? ? ? ?Arnd
>>
>> _______________________________________________
>> linaro-dev mailing list
>> linaro-dev at lists.linaro.org
>> http://lists.linaro.org/mailman/listinfo/linaro-dev
>
>
^ permalink raw reply [flat|nested] 65+ messages in thread
* [PATCH] i.MX23/28 framebuffer driver
2011-02-17 1:08 ` Clark, Rob
(?)
@ 2011-02-17 1:24 ` Jammy Zhou
2011-02-17 15:25 ` Clark, Rob
-1 siblings, 1 reply; 65+ messages in thread
From: Jammy Zhou @ 2011-02-17 1:24 UTC (permalink / raw)
To: linux-arm-kernel
There is already one KMS abstraction layer (libkms.so) in libdrm, maybe it
can serve as what we needed.
Regards,
Jammy
On Thu, Feb 17, 2011 at 9:08 AM, Clark, Rob <rob@ti.com> wrote:
> I'm still in the learning-as-I-go phase here, so definitely not ready
> to propose a solution, but it does seem to me like there is room for
> some sort of kms-helper library here to handle more of the boilerplate
> xf86-video-* stuff.. I guess I'll have a better picture once I have a
> chance to add support for the various multi-monitor configurations.
> But certainly would be interested if anyone already has some ideas.
>
> BR,
> -R
>
> On Wed, Feb 16, 2011 at 8:42 AM, Jesse Barker <jesse.barker@linaro.org>
> wrote:
> > Speaking for the Linaro graphics working group, I think it's great. And,
> I
> > think you're right, that if enough of the KMS support in xf86-video-* is
> > similar enough (I was only aware of intel and nouveau supporting it
> properly
> > at current), pulling it out into a common layer would make it easier to
> > support in new drivers (including fbdev).
> >
> > cheers,
> > Jesse
> >
> > On Wed, Feb 16, 2011 at 4:22 AM, Arnd Bergmann <arnd@arndb.de> wrote:
> >>
> >> On Tuesday 15 February 2011, Clark, Rob wrote:
> >> > I'd been experimenting a bit on the side w/ the DRM driver framework (
> >> >
> >> >
> http://gitorious.com/~robclark/pandaboard/robclarks-kernel-omap4/commits/omap_gpu
> >> > ), but had to add a good chunk of mostly boilerplate code to our xorg
> >> > driver in order just to test it. Maybe some generic support for KMS
> >> > in xf86-video-fbdev would have made this easier to develop the kernel
> >> > part without in parallel having to implement the userspace part. I'm
> >> > not sure if this is the sort of thing the linaro-wg has in mind?
> >>
> >> I'm not sure what the the linaro multimedia wg thinks of this, but the
> >> kernel code you linked looks like it's doing exactly the right thing.
> >>
> >> Arnd
> >>
> >> _______________________________________________
> >> linaro-dev mailing list
> >> linaro-dev at lists.linaro.org
> >> http://lists.linaro.org/mailman/listinfo/linaro-dev
> >
> >
>
> _______________________________________________
> linaro-dev mailing list
> linaro-dev at lists.linaro.org
> http://lists.linaro.org/mailman/listinfo/linaro-dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20110217/47c57980/attachment-0001.html>
^ permalink raw reply [flat|nested] 65+ messages in thread
* Re: [PATCH] i.MX23/28 framebuffer driver
2011-02-17 1:24 ` Jammy Zhou
@ 2011-02-17 15:25 ` Clark, Rob
0 siblings, 0 replies; 65+ messages in thread
From: Clark, Rob @ 2011-02-17 15:25 UTC (permalink / raw)
To: linux-arm-kernel
Hmm, I was thinking more on the xf86 side of things.. ie. to provide
default implementations of xf86CrtcFuncsRec and xf86OutputFuncsRec
functions..
BR,
-R
On Wed, Feb 16, 2011 at 7:24 PM, Jammy Zhou <jammy.zhou@linaro.org> wrote:
>
> There is already one KMS abstraction layer (libkms.so) in libdrm, maybe it can serve as what we needed.
>
> Regards,
> Jammy
>
> On Thu, Feb 17, 2011 at 9:08 AM, Clark, Rob <rob@ti.com> wrote:
>>
>> I'm still in the learning-as-I-go phase here, so definitely not ready
>> to propose a solution, but it does seem to me like there is room for
>> some sort of kms-helper library here to handle more of the boilerplate
>> xf86-video-* stuff.. I guess I'll have a better picture once I have a
>> chance to add support for the various multi-monitor configurations.
>> But certainly would be interested if anyone already has some ideas.
>>
>> BR,
>> -R
>>
>> On Wed, Feb 16, 2011 at 8:42 AM, Jesse Barker <jesse.barker@linaro.org> wrote:
>> > Speaking for the Linaro graphics working group, I think it's great. And, I
>> > think you're right, that if enough of the KMS support in xf86-video-* is
>> > similar enough (I was only aware of intel and nouveau supporting it properly
>> > at current), pulling it out into a common layer would make it easier to
>> > support in new drivers (including fbdev).
>> >
>> > cheers,
>> > Jesse
>> >
>> > On Wed, Feb 16, 2011 at 4:22 AM, Arnd Bergmann <arnd@arndb.de> wrote:
>> >>
>> >> On Tuesday 15 February 2011, Clark, Rob wrote:
>> >> > I'd been experimenting a bit on the side w/ the DRM driver framework (
>> >> >
>> >> > http://gitorious.com/~robclark/pandaboard/robclarks-kernel-omap4/commits/omap_gpu
>> >> > ), but had to add a good chunk of mostly boilerplate code to our xorg
>> >> > driver in order just to test it. Maybe some generic support for KMS
>> >> > in xf86-video-fbdev would have made this easier to develop the kernel
>> >> > part without in parallel having to implement the userspace part. I'm
>> >> > not sure if this is the sort of thing the linaro-wg has in mind?
>> >>
>> >> I'm not sure what the the linaro multimedia wg thinks of this, but the
>> >> kernel code you linked looks like it's doing exactly the right thing.
>> >>
>> >> Arnd
>> >>
>> >> _______________________________________________
>> >> linaro-dev mailing list
>> >> linaro-dev@lists.linaro.org
>> >> http://lists.linaro.org/mailman/listinfo/linaro-dev
>> >
>> >
>>
>> _______________________________________________
>> linaro-dev mailing list
>> linaro-dev@lists.linaro.org
>> http://lists.linaro.org/mailman/listinfo/linaro-dev
>
^ permalink raw reply [flat|nested] 65+ messages in thread
* [PATCH] i.MX23/28 framebuffer driver
@ 2011-02-17 15:25 ` Clark, Rob
0 siblings, 0 replies; 65+ messages in thread
From: Clark, Rob @ 2011-02-17 15:25 UTC (permalink / raw)
To: linux-arm-kernel
Hmm, I was thinking more on the xf86 side of things.. ie. to provide
default implementations of xf86CrtcFuncsRec and xf86OutputFuncsRec
functions..
BR,
-R
On Wed, Feb 16, 2011 at 7:24 PM, Jammy Zhou <jammy.zhou@linaro.org> wrote:
>
> There is already one KMS abstraction layer (libkms.so) in libdrm, maybe it can serve as what we needed.
>
> Regards,
> Jammy
>
> On Thu, Feb 17, 2011 at 9:08 AM, Clark, Rob <rob@ti.com> wrote:
>>
>> I'm still in the learning-as-I-go phase here, so definitely not ready
>> to propose a solution, but it does seem to me like there is room for
>> some sort of kms-helper library here to handle more of the boilerplate
>> xf86-video-* stuff.. ?I guess I'll have a better picture once I have a
>> chance to add support for the various multi-monitor configurations.
>> But certainly would be interested if anyone already has some ideas.
>>
>> BR,
>> -R
>>
>> On Wed, Feb 16, 2011 at 8:42 AM, Jesse Barker <jesse.barker@linaro.org> wrote:
>> > Speaking for the Linaro graphics working group, I think it's great.? And, I
>> > think you're right, that if enough of the KMS support in xf86-video-* is
>> > similar enough (I was only aware of intel and nouveau supporting it properly
>> > at current), pulling it out into a common layer would make it easier to
>> > support in new drivers (including fbdev).
>> >
>> > cheers,
>> > Jesse
>> >
>> > On Wed, Feb 16, 2011 at 4:22 AM, Arnd Bergmann <arnd@arndb.de> wrote:
>> >>
>> >> On Tuesday 15 February 2011, Clark, Rob wrote:
>> >> > I'd been experimenting a bit on the side w/ the DRM driver framework (
>> >> >
>> >> > http://gitorious.com/~robclark/pandaboard/robclarks-kernel-omap4/commits/omap_gpu
>> >> > ), but had to add a good chunk of mostly boilerplate code to our xorg
>> >> > driver in order just to test it. ?Maybe some generic support for KMS
>> >> > in xf86-video-fbdev would have made this easier to develop the kernel
>> >> > part without in parallel having to implement the userspace part. ?I'm
>> >> > not sure if this is the sort of thing the linaro-wg has in mind?
>> >>
>> >> I'm not sure what the the linaro multimedia wg thinks of this, but the
>> >> kernel code you linked looks like it's doing exactly the right thing.
>> >>
>> >> ? ? ? ?Arnd
>> >>
>> >> _______________________________________________
>> >> linaro-dev mailing list
>> >> linaro-dev at lists.linaro.org
>> >> http://lists.linaro.org/mailman/listinfo/linaro-dev
>> >
>> >
>>
>> _______________________________________________
>> linaro-dev mailing list
>> linaro-dev at lists.linaro.org
>> http://lists.linaro.org/mailman/listinfo/linaro-dev
>
^ permalink raw reply [flat|nested] 65+ messages in thread
* Re: [PATCH] i.MX23/28 framebuffer driver
2011-02-17 1:08 ` Clark, Rob
(?)
@ 2011-02-17 16:14 ` James Simmons
-1 siblings, 0 replies; 65+ messages in thread
From: James Simmons @ 2011-02-17 16:14 UTC (permalink / raw)
To: Clark, Rob
Cc: Linux Fbdev development list, linaro-dev, Arnd Bergmann,
Sascha Hauer, DRI development list, Jesse Barker,
Jakob Bornecrantz, linux-arm-kernel
> I'm still in the learning-as-I-go phase here, so definitely not ready
> to propose a solution, but it does seem to me like there is room for
> some sort of kms-helper library here to handle more of the boilerplate
> xf86-video-* stuff.. I guess I'll have a better picture once I have a
> chance to add support for the various multi-monitor configurations.
> But certainly would be interested if anyone already has some ideas.
I have been thinking about this as well. One of the short coming for DRM
on embedded platforms is the lack of a small well defined library that one
could use. Right now libkms only handles buffer allocation. The mode
setting is currently tied to the libdrm library. It would be nice to
seperate the two out more. Move the mode setting code to libkms and then
have libdrm be a wrapper around this. This way libdrm can both support
the legacy drm drivers and the new kms drivers at the same time. It also
makes a clear seperation. Jakob if you are willing to take this work I
will gladly seen you patches.
^ permalink raw reply [flat|nested] 65+ messages in thread
* [PATCH] i.MX23/28 framebuffer driver
@ 2011-02-17 16:14 ` James Simmons
0 siblings, 0 replies; 65+ messages in thread
From: James Simmons @ 2011-02-17 16:14 UTC (permalink / raw)
To: linux-arm-kernel
> I'm still in the learning-as-I-go phase here, so definitely not ready
> to propose a solution, but it does seem to me like there is room for
> some sort of kms-helper library here to handle more of the boilerplate
> xf86-video-* stuff.. I guess I'll have a better picture once I have a
> chance to add support for the various multi-monitor configurations.
> But certainly would be interested if anyone already has some ideas.
I have been thinking about this as well. One of the short coming for DRM
on embedded platforms is the lack of a small well defined library that one
could use. Right now libkms only handles buffer allocation. The mode
setting is currently tied to the libdrm library. It would be nice to
seperate the two out more. Move the mode setting code to libkms and then
have libdrm be a wrapper around this. This way libdrm can both support
the legacy drm drivers and the new kms drivers at the same time. It also
makes a clear seperation. Jakob if you are willing to take this work I
will gladly seen you patches.
^ permalink raw reply [flat|nested] 65+ messages in thread
* Re: [PATCH] i.MX23/28 framebuffer driver
@ 2011-02-17 16:14 ` James Simmons
0 siblings, 0 replies; 65+ messages in thread
From: James Simmons @ 2011-02-17 16:14 UTC (permalink / raw)
To: Clark, Rob
Cc: Linux Fbdev development list, linaro-dev, Arnd Bergmann,
Sascha Hauer, DRI development list, Jesse Barker,
Jakob Bornecrantz, linux-arm-kernel
> I'm still in the learning-as-I-go phase here, so definitely not ready
> to propose a solution, but it does seem to me like there is room for
> some sort of kms-helper library here to handle more of the boilerplate
> xf86-video-* stuff.. I guess I'll have a better picture once I have a
> chance to add support for the various multi-monitor configurations.
> But certainly would be interested if anyone already has some ideas.
I have been thinking about this as well. One of the short coming for DRM
on embedded platforms is the lack of a small well defined library that one
could use. Right now libkms only handles buffer allocation. The mode
setting is currently tied to the libdrm library. It would be nice to
seperate the two out more. Move the mode setting code to libkms and then
have libdrm be a wrapper around this. This way libdrm can both support
the legacy drm drivers and the new kms drivers at the same time. It also
makes a clear seperation. Jakob if you are willing to take this work I
will gladly seen you patches.
^ permalink raw reply [flat|nested] 65+ messages in thread
* [PATCH] i.MX23/28 framebuffer driver
@ 2011-02-18 2:19 ` Jammy Zhou
0 siblings, 0 replies; 65+ messages in thread
From: Jammy Zhou @ 2011-02-18 2:19 UTC (permalink / raw)
To: linux-arm-kernel
To accommodate the fact of independent display controller and GPU components
in ARM SOC, it will be better if we can separate KMS from DRM both in kernel
space and user space (i.e, create a new drivers/video/kms directory for
kernel side, move KMS related code in libdrm to libkms in user space). Then
we can build xrandr1.2+ support based on KMS for ARM platforms, even if DRM
is still not supported. Besides, for buffer management part (GEM) in DRM, if
possible, we can also make it an independent module leaving DRM just a
wrapper, so that the GEM stuff can be used more flexibly.
Regards,
Jammy
On Fri, Feb 18, 2011 at 12:14 AM, James Simmons <jsimmons@infradead.org>wrote:
>
> > I'm still in the learning-as-I-go phase here, so definitely not ready
> > to propose a solution, but it does seem to me like there is room for
> > some sort of kms-helper library here to handle more of the boilerplate
> > xf86-video-* stuff.. I guess I'll have a better picture once I have a
> > chance to add support for the various multi-monitor configurations.
> > But certainly would be interested if anyone already has some ideas.
>
> I have been thinking about this as well. One of the short coming for DRM
> on embedded platforms is the lack of a small well defined library that one
> could use. Right now libkms only handles buffer allocation. The mode
> setting is currently tied to the libdrm library. It would be nice to
> seperate the two out more. Move the mode setting code to libkms and then
> have libdrm be a wrapper around this. This way libdrm can both support
> the legacy drm drivers and the new kms drivers at the same time. It also
> makes a clear seperation. Jakob if you are willing to take this work I
> will gladly seen you patches.
>
> _______________________________________________
> linaro-dev mailing list
> linaro-dev at lists.linaro.org
> http://lists.linaro.org/mailman/listinfo/linaro-dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20110218/312cbc88/attachment-0001.html>
^ permalink raw reply [flat|nested] 65+ messages in thread
* Re: [PATCH] i.MX23/28 framebuffer driver
@ 2011-02-18 2:19 ` Jammy Zhou
0 siblings, 0 replies; 65+ messages in thread
From: Jammy Zhou @ 2011-02-18 2:19 UTC (permalink / raw)
To: James Simmons
Cc: Linux Fbdev development list, linaro-dev-cunTk1MwBs8s++Sfvej+rw,
Arnd Bergmann, Sascha Hauer, DRI development list,
Jakob Bornecrantz,
linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r
[-- Attachment #1.1: Type: text/plain, Size: 1961 bytes --]
To accommodate the fact of independent display controller and GPU components
in ARM SOC, it will be better if we can separate KMS from DRM both in kernel
space and user space (i.e, create a new drivers/video/kms directory for
kernel side, move KMS related code in libdrm to libkms in user space). Then
we can build xrandr1.2+ support based on KMS for ARM platforms, even if DRM
is still not supported. Besides, for buffer management part (GEM) in DRM, if
possible, we can also make it an independent module leaving DRM just a
wrapper, so that the GEM stuff can be used more flexibly.
Regards,
Jammy
On Fri, Feb 18, 2011 at 12:14 AM, James Simmons <jsimmons-wEGCiKHe2LqWVfeAwA7xHQ@public.gmane.org>wrote:
>
> > I'm still in the learning-as-I-go phase here, so definitely not ready
> > to propose a solution, but it does seem to me like there is room for
> > some sort of kms-helper library here to handle more of the boilerplate
> > xf86-video-* stuff.. I guess I'll have a better picture once I have a
> > chance to add support for the various multi-monitor configurations.
> > But certainly would be interested if anyone already has some ideas.
>
> I have been thinking about this as well. One of the short coming for DRM
> on embedded platforms is the lack of a small well defined library that one
> could use. Right now libkms only handles buffer allocation. The mode
> setting is currently tied to the libdrm library. It would be nice to
> seperate the two out more. Move the mode setting code to libkms and then
> have libdrm be a wrapper around this. This way libdrm can both support
> the legacy drm drivers and the new kms drivers at the same time. It also
> makes a clear seperation. Jakob if you are willing to take this work I
> will gladly seen you patches.
>
> _______________________________________________
> linaro-dev mailing list
> linaro-dev-cunTk1MwBs8s++Sfvej+rw@public.gmane.org
> http://lists.linaro.org/mailman/listinfo/linaro-dev
>
[-- Attachment #1.2: Type: text/html, Size: 2574 bytes --]
[-- Attachment #2: Type: text/plain, Size: 175 bytes --]
_______________________________________________
linaro-dev mailing list
linaro-dev-cunTk1MwBs8s++Sfvej+rw@public.gmane.org
http://lists.linaro.org/mailman/listinfo/linaro-dev
^ permalink raw reply [flat|nested] 65+ messages in thread
* [PATCH] i.MX23/28 framebuffer driver
2011-02-17 15:25 ` Clark, Rob
(?)
@ 2011-02-18 2:25 ` Jammy Zhou
2011-02-18 4:34 ` Clark, Rob
-1 siblings, 1 reply; 65+ messages in thread
From: Jammy Zhou @ 2011-02-18 2:25 UTC (permalink / raw)
To: linux-arm-kernel
I also noticed that default XRandR1.2+ implementation is missing in X side.
If we can implement one, it would be beneficial for all ARM platforms. By
the way, does X driver of TI support XRandR1.2+?
Regards,
Jammy
On Thu, Feb 17, 2011 at 11:25 PM, Clark, Rob <rob@ti.com> wrote:
> Hmm, I was thinking more on the xf86 side of things.. ie. to provide
> default implementations of xf86CrtcFuncsRec and xf86OutputFuncsRec
> functions..
>
> BR,
> -R
>
> On Wed, Feb 16, 2011 at 7:24 PM, Jammy Zhou <jammy.zhou@linaro.org> wrote:
> >
> > There is already one KMS abstraction layer (libkms.so) in libdrm, maybe
> it can serve as what we needed.
> >
> > Regards,
> > Jammy
> >
> > On Thu, Feb 17, 2011 at 9:08 AM, Clark, Rob <rob@ti.com> wrote:
> >>
> >> I'm still in the learning-as-I-go phase here, so definitely not ready
> >> to propose a solution, but it does seem to me like there is room for
> >> some sort of kms-helper library here to handle more of the boilerplate
> >> xf86-video-* stuff.. I guess I'll have a better picture once I have a
> >> chance to add support for the various multi-monitor configurations.
> >> But certainly would be interested if anyone already has some ideas.
> >>
> >> BR,
> >> -R
> >>
> >> On Wed, Feb 16, 2011 at 8:42 AM, Jesse Barker <jesse.barker@linaro.org>
> wrote:
> >> > Speaking for the Linaro graphics working group, I think it's great.
> And, I
> >> > think you're right, that if enough of the KMS support in xf86-video-*
> is
> >> > similar enough (I was only aware of intel and nouveau supporting it
> properly
> >> > at current), pulling it out into a common layer would make it easier
> to
> >> > support in new drivers (including fbdev).
> >> >
> >> > cheers,
> >> > Jesse
> >> >
> >> > On Wed, Feb 16, 2011 at 4:22 AM, Arnd Bergmann <arnd@arndb.de> wrote:
> >> >>
> >> >> On Tuesday 15 February 2011, Clark, Rob wrote:
> >> >> > I'd been experimenting a bit on the side w/ the DRM driver
> framework (
> >> >> >
> >> >> >
> http://gitorious.com/~robclark/pandaboard/robclarks-kernel-omap4/commits/omap_gpu
> >> >> > ), but had to add a good chunk of mostly boilerplate code to our
> xorg
> >> >> > driver in order just to test it. Maybe some generic support for
> KMS
> >> >> > in xf86-video-fbdev would have made this easier to develop the
> kernel
> >> >> > part without in parallel having to implement the userspace part.
> I'm
> >> >> > not sure if this is the sort of thing the linaro-wg has in mind?
> >> >>
> >> >> I'm not sure what the the linaro multimedia wg thinks of this, but
> the
> >> >> kernel code you linked looks like it's doing exactly the right thing.
> >> >>
> >> >> Arnd
> >> >>
> >> >> _______________________________________________
> >> >> linaro-dev mailing list
> >> >> linaro-dev at lists.linaro.org
> >> >> http://lists.linaro.org/mailman/listinfo/linaro-dev
> >> >
> >> >
> >>
> >> _______________________________________________
> >> linaro-dev mailing list
> >> linaro-dev at lists.linaro.org
> >> http://lists.linaro.org/mailman/listinfo/linaro-dev
> >
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20110218/dae57eaa/attachment-0001.html>
^ permalink raw reply [flat|nested] 65+ messages in thread
* Re: [PATCH] i.MX23/28 framebuffer driver
2011-02-18 2:19 ` Jammy Zhou
(?)
@ 2011-02-18 4:32 ` Clark, Rob
-1 siblings, 0 replies; 65+ messages in thread
From: Clark, Rob @ 2011-02-18 4:32 UTC (permalink / raw)
To: Jammy Zhou
Cc: Linux Fbdev development list, linaro-dev-cunTk1MwBs8s++Sfvej+rw,
Arnd Bergmann, Sascha Hauer, DRI development list, James Simmons,
Jakob Bornecrantz,
linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r
I'm all for a more modular drm, although I think the framework of
CRTCs, encoders, and connectors is a nice fit, at least for our hw.
It would be nice to have a better way to expose overlays. And I'm
still thinking about how/if GEM fits in with our various video and
2/3d accelerators.
BR,
-R
On Thu, Feb 17, 2011 at 8:19 PM, Jammy Zhou <jammy.zhou@linaro.org> wrote:
> To accommodate the fact of independent display controller and GPU components
> in ARM SOC, it will be better if we can separate KMS from DRM both in kernel
> space and user space (i.e, create a new drivers/video/kms directory for
> kernel side, move KMS related code in libdrm to libkms in user space). Then
> we can build xrandr1.2+ support based on KMS for ARM platforms, even if DRM
> is still not supported. Besides, for buffer management part (GEM) in DRM, if
> possible, we can also make it an independent module leaving DRM just a
> wrapper, so that the GEM stuff can be used more flexibly.
>
> Regards,
> Jammy
>
> On Fri, Feb 18, 2011 at 12:14 AM, James Simmons <jsimmons@infradead.org>
> wrote:
>>
>> > I'm still in the learning-as-I-go phase here, so definitely not ready
>> > to propose a solution, but it does seem to me like there is room for
>> > some sort of kms-helper library here to handle more of the boilerplate
>> > xf86-video-* stuff.. I guess I'll have a better picture once I have a
>> > chance to add support for the various multi-monitor configurations.
>> > But certainly would be interested if anyone already has some ideas.
>>
>> I have been thinking about this as well. One of the short coming for DRM
>> on embedded platforms is the lack of a small well defined library that one
>> could use. Right now libkms only handles buffer allocation. The mode
>> setting is currently tied to the libdrm library. It would be nice to
>> seperate the two out more. Move the mode setting code to libkms and then
>> have libdrm be a wrapper around this. This way libdrm can both support
>> the legacy drm drivers and the new kms drivers at the same time. It also
>> makes a clear seperation. Jakob if you are willing to take this work I
>> will gladly seen you patches.
>>
>> _______________________________________________
>> linaro-dev mailing list
>> linaro-dev@lists.linaro.org
>> http://lists.linaro.org/mailman/listinfo/linaro-dev
>
>
^ permalink raw reply [flat|nested] 65+ messages in thread
* [PATCH] i.MX23/28 framebuffer driver
@ 2011-02-18 4:32 ` Clark, Rob
0 siblings, 0 replies; 65+ messages in thread
From: Clark, Rob @ 2011-02-18 4:32 UTC (permalink / raw)
To: linux-arm-kernel
I'm all for a more modular drm, although I think the framework of
CRTCs, encoders, and connectors is a nice fit, at least for our hw.
It would be nice to have a better way to expose overlays. And I'm
still thinking about how/if GEM fits in with our various video and
2/3d accelerators.
BR,
-R
On Thu, Feb 17, 2011 at 8:19 PM, Jammy Zhou <jammy.zhou@linaro.org> wrote:
> To accommodate the fact of independent display controller and GPU components
> in ARM SOC, it will be better if we can separate KMS from DRM both in kernel
> space and user space (i.e, create a new drivers/video/kms directory for
> kernel side, move KMS related code in libdrm to libkms in user space). Then
> we can build xrandr1.2+ support based on KMS for ARM platforms, even if DRM
> is still not supported. Besides, for buffer management part (GEM) in DRM, if
> possible, we can also make it an independent module leaving DRM just a
> wrapper, so that the GEM stuff can be used more flexibly.
>
> Regards,
> Jammy
>
> On Fri, Feb 18, 2011 at 12:14 AM, James Simmons <jsimmons@infradead.org>
> wrote:
>>
>> > I'm still in the learning-as-I-go phase here, so definitely not ready
>> > to propose a solution, but it does seem to me like there is room for
>> > some sort of kms-helper library here to handle more of the boilerplate
>> > xf86-video-* stuff.. ?I guess I'll have a better picture once I have a
>> > chance to add support for the various multi-monitor configurations.
>> > But certainly would be interested if anyone already has some ideas.
>>
>> I have been thinking about this as well. One of the short coming for DRM
>> on embedded platforms is the lack of a small well defined library that one
>> could use. Right now libkms only handles buffer allocation. The mode
>> setting is currently tied to the libdrm library. It would be nice to
>> seperate the two out more. Move the mode setting code to libkms and then
>> have libdrm be a wrapper around this. This way libdrm can both support
>> the legacy drm drivers and the new kms drivers at the same time. It also
>> makes a clear seperation. Jakob if you are willing to take this work I
>> will gladly seen you patches.
>>
>> _______________________________________________
>> linaro-dev mailing list
>> linaro-dev at lists.linaro.org
>> http://lists.linaro.org/mailman/listinfo/linaro-dev
>
>
^ permalink raw reply [flat|nested] 65+ messages in thread
* Re: [PATCH] i.MX23/28 framebuffer driver
@ 2011-02-18 4:32 ` Clark, Rob
0 siblings, 0 replies; 65+ messages in thread
From: Clark, Rob @ 2011-02-18 4:32 UTC (permalink / raw)
To: Jammy Zhou
Cc: Linux Fbdev development list, linaro-dev-cunTk1MwBs8s++Sfvej+rw,
Arnd Bergmann, Sascha Hauer, DRI development list, James Simmons,
Jakob Bornecrantz,
linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r
I'm all for a more modular drm, although I think the framework of
CRTCs, encoders, and connectors is a nice fit, at least for our hw.
It would be nice to have a better way to expose overlays. And I'm
still thinking about how/if GEM fits in with our various video and
2/3d accelerators.
BR,
-R
On Thu, Feb 17, 2011 at 8:19 PM, Jammy Zhou <jammy.zhou-QSEj5FYQhm4dnm+yROfE0A@public.gmane.org> wrote:
> To accommodate the fact of independent display controller and GPU components
> in ARM SOC, it will be better if we can separate KMS from DRM both in kernel
> space and user space (i.e, create a new drivers/video/kms directory for
> kernel side, move KMS related code in libdrm to libkms in user space). Then
> we can build xrandr1.2+ support based on KMS for ARM platforms, even if DRM
> is still not supported. Besides, for buffer management part (GEM) in DRM, if
> possible, we can also make it an independent module leaving DRM just a
> wrapper, so that the GEM stuff can be used more flexibly.
>
> Regards,
> Jammy
>
> On Fri, Feb 18, 2011 at 12:14 AM, James Simmons <jsimmons-wEGCiKHe2LqWVfeAwA7xHQ@public.gmane.org>
> wrote:
>>
>> > I'm still in the learning-as-I-go phase here, so definitely not ready
>> > to propose a solution, but it does seem to me like there is room for
>> > some sort of kms-helper library here to handle more of the boilerplate
>> > xf86-video-* stuff.. I guess I'll have a better picture once I have a
>> > chance to add support for the various multi-monitor configurations.
>> > But certainly would be interested if anyone already has some ideas.
>>
>> I have been thinking about this as well. One of the short coming for DRM
>> on embedded platforms is the lack of a small well defined library that one
>> could use. Right now libkms only handles buffer allocation. The mode
>> setting is currently tied to the libdrm library. It would be nice to
>> seperate the two out more. Move the mode setting code to libkms and then
>> have libdrm be a wrapper around this. This way libdrm can both support
>> the legacy drm drivers and the new kms drivers at the same time. It also
>> makes a clear seperation. Jakob if you are willing to take this work I
>> will gladly seen you patches.
>>
>> _______________________________________________
>> linaro-dev mailing list
>> linaro-dev-cunTk1MwBs8s++Sfvej+rw@public.gmane.org
>> http://lists.linaro.org/mailman/listinfo/linaro-dev
>
>
^ permalink raw reply [flat|nested] 65+ messages in thread
* Re: [PATCH] i.MX23/28 framebuffer driver
2011-02-18 2:25 ` Jammy Zhou
@ 2011-02-18 4:34 ` Clark, Rob
0 siblings, 0 replies; 65+ messages in thread
From: Clark, Rob @ 2011-02-18 4:34 UTC (permalink / raw)
To: linux-arm-kernel
I'm in the process of adding xrandr support to our xorg driver..
definitely more built-in support on the X side would make this easier.
BR,
-R
On Thu, Feb 17, 2011 at 8:25 PM, Jammy Zhou <jammy.zhou@linaro.org> wrote:
> I also noticed that default XRandR1.2+ implementation is missing in X side.
> If we can implement one, it would be beneficial for all ARM platforms. By
> the way, does X driver of TI support XRandR1.2+?
>
> Regards,
> Jammy
>
> On Thu, Feb 17, 2011 at 11:25 PM, Clark, Rob <rob@ti.com> wrote:
>>
>> Hmm, I was thinking more on the xf86 side of things.. ie. to provide
>> default implementations of xf86CrtcFuncsRec and xf86OutputFuncsRec
>> functions..
>>
>> BR,
>> -R
>>
>> On Wed, Feb 16, 2011 at 7:24 PM, Jammy Zhou <jammy.zhou@linaro.org> wrote:
>> >
>> > There is already one KMS abstraction layer (libkms.so) in libdrm, maybe
>> > it can serve as what we needed.
>> >
>> > Regards,
>> > Jammy
>> >
>> > On Thu, Feb 17, 2011 at 9:08 AM, Clark, Rob <rob@ti.com> wrote:
>> >>
>> >> I'm still in the learning-as-I-go phase here, so definitely not ready
>> >> to propose a solution, but it does seem to me like there is room for
>> >> some sort of kms-helper library here to handle more of the boilerplate
>> >> xf86-video-* stuff.. I guess I'll have a better picture once I have a
>> >> chance to add support for the various multi-monitor configurations.
>> >> But certainly would be interested if anyone already has some ideas.
>> >>
>> >> BR,
>> >> -R
>> >>
>> >> On Wed, Feb 16, 2011 at 8:42 AM, Jesse Barker <jesse.barker@linaro.org>
>> >> wrote:
>> >> > Speaking for the Linaro graphics working group, I think it's great.
>> >> > And, I
>> >> > think you're right, that if enough of the KMS support in xf86-video-*
>> >> > is
>> >> > similar enough (I was only aware of intel and nouveau supporting it
>> >> > properly
>> >> > at current), pulling it out into a common layer would make it easier
>> >> > to
>> >> > support in new drivers (including fbdev).
>> >> >
>> >> > cheers,
>> >> > Jesse
>> >> >
>> >> > On Wed, Feb 16, 2011 at 4:22 AM, Arnd Bergmann <arnd@arndb.de> wrote:
>> >> >>
>> >> >> On Tuesday 15 February 2011, Clark, Rob wrote:
>> >> >> > I'd been experimenting a bit on the side w/ the DRM driver
>> >> >> > framework (
>> >> >> >
>> >> >> >
>> >> >> > http://gitorious.com/~robclark/pandaboard/robclarks-kernel-omap4/commits/omap_gpu
>> >> >> > ), but had to add a good chunk of mostly boilerplate code to our
>> >> >> > xorg
>> >> >> > driver in order just to test it. Maybe some generic support for
>> >> >> > KMS
>> >> >> > in xf86-video-fbdev would have made this easier to develop the
>> >> >> > kernel
>> >> >> > part without in parallel having to implement the userspace part.
>> >> >> > I'm
>> >> >> > not sure if this is the sort of thing the linaro-wg has in mind?
>> >> >>
>> >> >> I'm not sure what the the linaro multimedia wg thinks of this, but
>> >> >> the
>> >> >> kernel code you linked looks like it's doing exactly the right
>> >> >> thing.
>> >> >>
>> >> >> Arnd
>> >> >>
>> >> >> _______________________________________________
>> >> >> linaro-dev mailing list
>> >> >> linaro-dev@lists.linaro.org
>> >> >> http://lists.linaro.org/mailman/listinfo/linaro-dev
>> >> >
>> >> >
>> >>
>> >> _______________________________________________
>> >> linaro-dev mailing list
>> >> linaro-dev@lists.linaro.org
>> >> http://lists.linaro.org/mailman/listinfo/linaro-dev
>> >
>
>
^ permalink raw reply [flat|nested] 65+ messages in thread
* [PATCH] i.MX23/28 framebuffer driver
@ 2011-02-18 4:34 ` Clark, Rob
0 siblings, 0 replies; 65+ messages in thread
From: Clark, Rob @ 2011-02-18 4:34 UTC (permalink / raw)
To: linux-arm-kernel
I'm in the process of adding xrandr support to our xorg driver..
definitely more built-in support on the X side would make this easier.
BR,
-R
On Thu, Feb 17, 2011 at 8:25 PM, Jammy Zhou <jammy.zhou@linaro.org> wrote:
> I also noticed that default XRandR1.2+ implementation is missing in X side.
> If we can implement one, it would be beneficial for all ARM platforms. By
> the way, does X driver of TI support XRandR1.2+?
>
> Regards,
> Jammy
>
> On Thu, Feb 17, 2011 at 11:25 PM, Clark, Rob <rob@ti.com> wrote:
>>
>> Hmm, I was thinking more on the xf86 side of things.. ie. to provide
>> default implementations of xf86CrtcFuncsRec and xf86OutputFuncsRec
>> functions..
>>
>> BR,
>> -R
>>
>> On Wed, Feb 16, 2011 at 7:24 PM, Jammy Zhou <jammy.zhou@linaro.org> wrote:
>> >
>> > There is already one KMS abstraction layer (libkms.so) in libdrm, maybe
>> > it can serve as what we needed.
>> >
>> > Regards,
>> > Jammy
>> >
>> > On Thu, Feb 17, 2011 at 9:08 AM, Clark, Rob <rob@ti.com> wrote:
>> >>
>> >> I'm still in the learning-as-I-go phase here, so definitely not ready
>> >> to propose a solution, but it does seem to me like there is room for
>> >> some sort of kms-helper library here to handle more of the boilerplate
>> >> xf86-video-* stuff.. ?I guess I'll have a better picture once I have a
>> >> chance to add support for the various multi-monitor configurations.
>> >> But certainly would be interested if anyone already has some ideas.
>> >>
>> >> BR,
>> >> -R
>> >>
>> >> On Wed, Feb 16, 2011 at 8:42 AM, Jesse Barker <jesse.barker@linaro.org>
>> >> wrote:
>> >> > Speaking for the Linaro graphics working group, I think it's great.
>> >> > And, I
>> >> > think you're right, that if enough of the KMS support in xf86-video-*
>> >> > is
>> >> > similar enough (I was only aware of intel and nouveau supporting it
>> >> > properly
>> >> > at current), pulling it out into a common layer would make it easier
>> >> > to
>> >> > support in new drivers (including fbdev).
>> >> >
>> >> > cheers,
>> >> > Jesse
>> >> >
>> >> > On Wed, Feb 16, 2011 at 4:22 AM, Arnd Bergmann <arnd@arndb.de> wrote:
>> >> >>
>> >> >> On Tuesday 15 February 2011, Clark, Rob wrote:
>> >> >> > I'd been experimenting a bit on the side w/ the DRM driver
>> >> >> > framework (
>> >> >> >
>> >> >> >
>> >> >> > http://gitorious.com/~robclark/pandaboard/robclarks-kernel-omap4/commits/omap_gpu
>> >> >> > ), but had to add a good chunk of mostly boilerplate code to our
>> >> >> > xorg
>> >> >> > driver in order just to test it. ?Maybe some generic support for
>> >> >> > KMS
>> >> >> > in xf86-video-fbdev would have made this easier to develop the
>> >> >> > kernel
>> >> >> > part without in parallel having to implement the userspace part.
>> >> >> > ?I'm
>> >> >> > not sure if this is the sort of thing the linaro-wg has in mind?
>> >> >>
>> >> >> I'm not sure what the the linaro multimedia wg thinks of this, but
>> >> >> the
>> >> >> kernel code you linked looks like it's doing exactly the right
>> >> >> thing.
>> >> >>
>> >> >> ? ? ? ?Arnd
>> >> >>
>> >> >> _______________________________________________
>> >> >> linaro-dev mailing list
>> >> >> linaro-dev at lists.linaro.org
>> >> >> http://lists.linaro.org/mailman/listinfo/linaro-dev
>> >> >
>> >> >
>> >>
>> >> _______________________________________________
>> >> linaro-dev mailing list
>> >> linaro-dev at lists.linaro.org
>> >> http://lists.linaro.org/mailman/listinfo/linaro-dev
>> >
>
>
^ permalink raw reply [flat|nested] 65+ messages in thread
* Re: [PATCH] i.MX23/28 framebuffer driver
2011-02-17 16:14 ` James Simmons
(?)
@ 2011-02-21 3:12 ` Dave Airlie
-1 siblings, 0 replies; 65+ messages in thread
From: Dave Airlie @ 2011-02-21 3:12 UTC (permalink / raw)
To: James Simmons
Cc: Linux Fbdev development list, linaro-dev, Arnd Bergmann,
Sascha Hauer, DRI development list, Jesse Barker, Clark, Rob,
linux-arm-kernel
On Fri, Feb 18, 2011 at 2:14 AM, James Simmons <jsimmons@infradead.org> wrote:
>
>> I'm still in the learning-as-I-go phase here, so definitely not ready
>> to propose a solution, but it does seem to me like there is room for
>> some sort of kms-helper library here to handle more of the boilerplate
>> xf86-video-* stuff.. I guess I'll have a better picture once I have a
>> chance to add support for the various multi-monitor configurations.
>> But certainly would be interested if anyone already has some ideas.
>
> I have been thinking about this as well. One of the short coming for DRM
> on embedded platforms is the lack of a small well defined library that one
> could use. Right now libkms only handles buffer allocation. The mode
> setting is currently tied to the libdrm library. It would be nice to
> seperate the two out more. Move the mode setting code to libkms and then
> have libdrm be a wrapper around this. This way libdrm can both support
> the legacy drm drivers and the new kms drivers at the same time. It also
> makes a clear seperation. Jakob if you are willing to take this work I
> will gladly seen you patches.
the problem with that is libkms relies on libdrm, so you'd have a
circular dependency.
With the addition of the dumb ioctl to the kernel we can certainly
avoid some of that dependency,
but you guys are missing one important point about libkms, its meant
to be a fallback.
Why?
well because most GPUs in reality have non-generic memory layouts,
they prefer having tiled buffers for different things and different
types of tiling. There is simply no nice way to abstract that out,
since it depends on the userspace libraries that use this. Starting
off with an untiled allocation for the frontbuffer can pretty much
leave you messed up when you plug anything like acceleration in
afterwards.
I have been looking at this from the USB driver pov as well, and we
can probably resurrect the old xf86-video-modesetting driver along
with some code in the X server maybe, mesa also has a libkms fallback
X driver that works quite well if just a bit messy to build (since its
in mesa).
Dave.
^ permalink raw reply [flat|nested] 65+ messages in thread
* [PATCH] i.MX23/28 framebuffer driver
@ 2011-02-21 3:12 ` Dave Airlie
0 siblings, 0 replies; 65+ messages in thread
From: Dave Airlie @ 2011-02-21 3:12 UTC (permalink / raw)
To: linux-arm-kernel
On Fri, Feb 18, 2011 at 2:14 AM, James Simmons <jsimmons@infradead.org> wrote:
>
>> I'm still in the learning-as-I-go phase here, so definitely not ready
>> to propose a solution, but it does seem to me like there is room for
>> some sort of kms-helper library here to handle more of the boilerplate
>> xf86-video-* stuff.. ?I guess I'll have a better picture once I have a
>> chance to add support for the various multi-monitor configurations.
>> But certainly would be interested if anyone already has some ideas.
>
> I have been thinking about this as well. One of the short coming for DRM
> on embedded platforms is the lack of a small well defined library that one
> could use. Right now libkms only handles buffer allocation. The mode
> setting is currently tied to the libdrm library. It would be nice to
> seperate the two out more. Move the mode setting code to libkms and then
> have libdrm be a wrapper around this. This way libdrm can both support
> the legacy drm drivers and the new kms drivers at the same time. It also
> makes a clear seperation. Jakob if you are willing to take this work I
> will gladly seen you patches.
the problem with that is libkms relies on libdrm, so you'd have a
circular dependency.
With the addition of the dumb ioctl to the kernel we can certainly
avoid some of that dependency,
but you guys are missing one important point about libkms, its meant
to be a fallback.
Why?
well because most GPUs in reality have non-generic memory layouts,
they prefer having tiled buffers for different things and different
types of tiling. There is simply no nice way to abstract that out,
since it depends on the userspace libraries that use this. Starting
off with an untiled allocation for the frontbuffer can pretty much
leave you messed up when you plug anything like acceleration in
afterwards.
I have been looking at this from the USB driver pov as well, and we
can probably resurrect the old xf86-video-modesetting driver along
with some code in the X server maybe, mesa also has a libkms fallback
X driver that works quite well if just a bit messy to build (since its
in mesa).
Dave.
^ permalink raw reply [flat|nested] 65+ messages in thread
* Re: [PATCH] i.MX23/28 framebuffer driver
@ 2011-02-21 3:12 ` Dave Airlie
0 siblings, 0 replies; 65+ messages in thread
From: Dave Airlie @ 2011-02-21 3:12 UTC (permalink / raw)
To: James Simmons
Cc: Linux Fbdev development list, linaro-dev, Arnd Bergmann,
Sascha Hauer, DRI development list, Jesse Barker, Clark, Rob,
linux-arm-kernel
On Fri, Feb 18, 2011 at 2:14 AM, James Simmons <jsimmons@infradead.org> wrote:
>
>> I'm still in the learning-as-I-go phase here, so definitely not ready
>> to propose a solution, but it does seem to me like there is room for
>> some sort of kms-helper library here to handle more of the boilerplate
>> xf86-video-* stuff.. I guess I'll have a better picture once I have a
>> chance to add support for the various multi-monitor configurations.
>> But certainly would be interested if anyone already has some ideas.
>
> I have been thinking about this as well. One of the short coming for DRM
> on embedded platforms is the lack of a small well defined library that one
> could use. Right now libkms only handles buffer allocation. The mode
> setting is currently tied to the libdrm library. It would be nice to
> seperate the two out more. Move the mode setting code to libkms and then
> have libdrm be a wrapper around this. This way libdrm can both support
> the legacy drm drivers and the new kms drivers at the same time. It also
> makes a clear seperation. Jakob if you are willing to take this work I
> will gladly seen you patches.
the problem with that is libkms relies on libdrm, so you'd have a
circular dependency.
With the addition of the dumb ioctl to the kernel we can certainly
avoid some of that dependency,
but you guys are missing one important point about libkms, its meant
to be a fallback.
Why?
well because most GPUs in reality have non-generic memory layouts,
they prefer having tiled buffers for different things and different
types of tiling. There is simply no nice way to abstract that out,
since it depends on the userspace libraries that use this. Starting
off with an untiled allocation for the frontbuffer can pretty much
leave you messed up when you plug anything like acceleration in
afterwards.
I have been looking at this from the USB driver pov as well, and we
can probably resurrect the old xf86-video-modesetting driver along
with some code in the X server maybe, mesa also has a libkms fallback
X driver that works quite well if just a bit messy to build (since its
in mesa).
Dave.
^ permalink raw reply [flat|nested] 65+ messages in thread
end of thread, other threads:[~2011-02-21 3:12 UTC | newest]
Thread overview: 65+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2011-02-09 13:20 [PATCH] i.MX23/28 framebuffer driver Sascha Hauer
2011-02-09 13:20 ` [PATCH 1/2] video: Add " Sascha Hauer
2011-02-09 13:20 ` Sascha Hauer
2011-02-09 13:35 ` Lothar Waßmann
2011-02-09 13:35 ` Lothar Waßmann
2011-02-10 3:16 ` Li Frank-B20596
2011-02-10 3:16 ` Li Frank-B20596
2011-02-10 8:51 ` Juergen Beisert
2011-02-10 8:51 ` Juergen Beisert
2011-02-10 9:15 ` Li Frank-B20596
2011-02-10 9:15 ` Li Frank-B20596
2011-02-10 9:52 ` Juergen Beisert
2011-02-10 9:52 ` Juergen Beisert
2011-02-10 10:37 ` Li Frank-B20596
2011-02-10 10:37 ` Li Frank-B20596
2011-02-10 11:12 ` Juergen Beisert
2011-02-10 11:12 ` Juergen Beisert
2011-02-10 12:23 ` Li Frank-B20596
2011-02-10 12:23 ` Li Frank-B20596
2011-02-10 9:23 ` Li Frank-B20596
2011-02-10 9:23 ` Li Frank-B20596
2011-02-10 9:46 ` Juergen Beisert
2011-02-10 9:46 ` Juergen Beisert
2011-02-10 10:08 ` Li Frank-B20596
2011-02-10 10:08 ` Li Frank-B20596
2011-02-10 10:47 ` Wolfram Sang
2011-02-10 10:47 ` Wolfram Sang
2011-02-10 11:51 ` Sascha Hauer
2011-02-10 11:51 ` Sascha Hauer
2011-02-10 12:32 ` Li Frank-B20596
2011-02-10 12:32 ` Li Frank-B20596
2011-02-09 13:20 ` [PATCH 2/2] ARM i.MX23/28: Add framebuffer device support Sascha Hauer
2011-02-09 14:33 ` Fabio Estevam
2011-02-10 3:23 ` Li Frank-B20596
2011-02-09 14:47 ` [PATCH] i.MX23/28 framebuffer driver Arnd Bergmann
2011-02-09 15:37 ` Sascha Hauer
2011-02-09 16:31 ` Arnd Bergmann
2011-02-09 16:31 ` Arnd Bergmann
2011-02-10 17:09 ` Robert Schwebel
2011-02-10 17:09 ` Robert Schwebel
2011-02-15 14:13 ` Clark, Rob
2011-02-15 14:13 ` Clark, Rob
2011-02-16 12:22 ` Arnd Bergmann
2011-02-16 12:22 ` Arnd Bergmann
2011-02-16 14:42 ` Jesse Barker
2011-02-17 1:08 ` Clark, Rob
2011-02-17 1:08 ` Clark, Rob
2011-02-17 1:24 ` Jammy Zhou
2011-02-17 15:25 ` Clark, Rob
2011-02-17 15:25 ` Clark, Rob
2011-02-18 2:25 ` Jammy Zhou
2011-02-18 4:34 ` Clark, Rob
2011-02-18 4:34 ` Clark, Rob
2011-02-17 16:14 ` James Simmons
2011-02-17 16:14 ` James Simmons
2011-02-17 16:14 ` James Simmons
2011-02-18 2:19 ` Jammy Zhou
2011-02-18 2:19 ` Jammy Zhou
[not found] ` <AANLkTimcwr4vpycSQCH9vMaB+umh+yHD_+WD2MJNqMOB-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
2011-02-18 4:32 ` Clark, Rob
2011-02-18 4:32 ` Clark, Rob
2011-02-18 4:32 ` Clark, Rob
2011-02-21 3:12 ` Dave Airlie
2011-02-21 3:12 ` Dave Airlie
2011-02-21 3:12 ` Dave Airlie
2011-02-10 9:22 ` Domenico Andreoli
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.