The Versatile Express uses a special configuration controller deeply embedded in the system motherboard FPGA to multiplex the two to three (!) display controller instances out to the single SiI9022 bridge.
Set up an extra file with the logic to probe to the FPGA mux register on the system controller bus, then parse the device tree to see if there is a CLCD or HDLCD instance on the core tile (also known as the daughterboard) by looking in the root of the device tree for compatible nodes.
- If there is a HDLCD on the core tile, and there is a driver for it, we exit probe and deactivate the motherboard CLCD. We do not touch the DVI mux in this case, to make sure we don't break HDLCD.
- If there is a CLCD on both the motherboard and the core tile (only the CA9 has this) the core tile CLCD takes precedence and get muxed to the DVI connector.
- Only if there is no working graphics on the core tile, the motherboard CLCD is probed and muxed to the DVI connector.
Core tile graphics should always take precedence as it can address all memory and is also faster, however the motherboard CLCD is good to have around for diagnostics and testing.
It is possible to test the motherboard CLCD by setting the status = "disabled" property on the core tile CLCD or HDLCD.
Scale down the Versatile Express to 16BPP so we can support a 1024x768 display despite the bus bandwidth restrictions on this platform. (The motherboard CLCD supports slightly lower resolution.)
Cc: Liviu Dudau liviu.dudau@arm.com Cc: Pawel Moll pawel.moll@arm.com Cc: Robin Murphy robin.murphy@arm.com Acked-by: Eric Anholt eric@anholt.net Signed-off-by: Linus Walleij linus.walleij@linaro.org --- ChangeLog v3->v4: - Add a missing platform_device_put() - Simplify code with for_each_available_child_of_node() - Collect Eric's ACK. ChangeLog v2->v3: - Rewrite CLCD detection and mux priority logic, look in the device tree root for core tile graphics. ChangeLog v1->v2: - No changes just reposting rebased on mainline changes. --- drivers/gpu/drm/pl111/Makefile | 1 + drivers/gpu/drm/pl111/pl111_versatile.c | 49 ++++++++++++- drivers/gpu/drm/pl111/pl111_vexpress.c | 125 ++++++++++++++++++++++++++++++++ drivers/gpu/drm/pl111/pl111_vexpress.h | 22 ++++++ 4 files changed, 196 insertions(+), 1 deletion(-) create mode 100644 drivers/gpu/drm/pl111/pl111_vexpress.c create mode 100644 drivers/gpu/drm/pl111/pl111_vexpress.h
diff --git a/drivers/gpu/drm/pl111/Makefile b/drivers/gpu/drm/pl111/Makefile index 9c5e8dba8ac6..19a8189dc54f 100644 --- a/drivers/gpu/drm/pl111/Makefile +++ b/drivers/gpu/drm/pl111/Makefile @@ -3,6 +3,7 @@ pl111_drm-y += pl111_display.o \ pl111_versatile.o \ pl111_drv.o
+pl111_drm-$(CONFIG_ARCH_VEXPRESS) += pl111_vexpress.o pl111_drm-$(CONFIG_DEBUG_FS) += pl111_debugfs.o
obj-$(CONFIG_DRM_PL111) += pl111_drm.o diff --git a/drivers/gpu/drm/pl111/pl111_versatile.c b/drivers/gpu/drm/pl111/pl111_versatile.c index 9302f516045e..78ddf8534fd2 100644 --- a/drivers/gpu/drm/pl111/pl111_versatile.c +++ b/drivers/gpu/drm/pl111/pl111_versatile.c @@ -1,12 +1,14 @@ #include <linux/amba/clcd-regs.h> #include <linux/device.h> #include <linux/of.h> +#include <linux/of_platform.h> #include <linux/regmap.h> #include <linux/mfd/syscon.h> #include <linux/bitops.h> #include <linux/module.h> #include <drm/drmP.h> #include "pl111_versatile.h" +#include "pl111_vexpress.h" #include "pl111_drm.h"
static struct regmap *versatile_syscon_map; @@ -22,6 +24,7 @@ enum versatile_clcd { REALVIEW_CLCD_PB11MP, REALVIEW_CLCD_PBA8, REALVIEW_CLCD_PBX, + VEXPRESS_CLCD_V2M, };
static const struct of_device_id versatile_clcd_of_match[] = { @@ -53,6 +56,10 @@ static const struct of_device_id versatile_clcd_of_match[] = { .compatible = "arm,realview-pbx-syscon", .data = (void *)REALVIEW_CLCD_PBX, }, + { + .compatible = "arm,vexpress-muxfpga", + .data = (void *)VEXPRESS_CLCD_V2M, + }, {}, };
@@ -286,12 +293,26 @@ static const struct pl111_variant_data pl111_realview = { .fb_bpp = 16, };
+/* + * Versatile Express PL111 variant, again we just push the maximum + * BPP to 16 to be able to get 1024x768 without saturating the memory + * bus. The clockdivider also seems broken on the Versatile Express. + */ +static const struct pl111_variant_data pl111_vexpress = { + .name = "PL111 Versatile Express", + .formats = pl111_realview_pixel_formats, + .nformats = ARRAY_SIZE(pl111_realview_pixel_formats), + .fb_bpp = 16, + .broken_clockdivider = true, +}; + int pl111_versatile_init(struct device *dev, struct pl111_drm_dev_private *priv) { const struct of_device_id *clcd_id; enum versatile_clcd versatile_clcd_type; struct device_node *np; struct regmap *map; + int ret;
np = of_find_matching_node_and_match(NULL, versatile_clcd_of_match, &clcd_id); @@ -301,7 +322,26 @@ int pl111_versatile_init(struct device *dev, struct pl111_drm_dev_private *priv) } versatile_clcd_type = (enum versatile_clcd)clcd_id->data;
- map = syscon_node_to_regmap(np); + /* Versatile Express special handling */ + if (versatile_clcd_type == VEXPRESS_CLCD_V2M) { + struct platform_device *pdev; + + /* Call into deep Vexpress configuration API */ + pdev = of_find_device_by_node(np); + if (!pdev) { + dev_err(dev, "can't find the sysreg device, deferring\n"); + return -EPROBE_DEFER; + } + map = dev_get_drvdata(&pdev->dev); + if (!map) { + dev_err(dev, "sysreg has not yet probed\n"); + platform_device_put(pdev); + return -EPROBE_DEFER; + } + } else { + map = syscon_node_to_regmap(np); + } + if (IS_ERR(map)) { dev_err(dev, "no Versatile syscon regmap\n"); return PTR_ERR(map); @@ -340,6 +380,13 @@ int pl111_versatile_init(struct device *dev, struct pl111_drm_dev_private *priv) priv->variant_display_disable = pl111_realview_clcd_disable; dev_info(dev, "set up callbacks for RealView PL111\n"); break; + case VEXPRESS_CLCD_V2M: + priv->variant = &pl111_vexpress; + dev_info(dev, "initializing Versatile Express PL111\n"); + ret = pl111_vexpress_clcd_init(dev, priv, map); + if (ret) + return ret; + break; default: dev_info(dev, "unknown Versatile system controller\n"); break; diff --git a/drivers/gpu/drm/pl111/pl111_vexpress.c b/drivers/gpu/drm/pl111/pl111_vexpress.c new file mode 100644 index 000000000000..9cdb28621ff4 --- /dev/null +++ b/drivers/gpu/drm/pl111/pl111_vexpress.c @@ -0,0 +1,125 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Versatile Express PL111 handling + * Copyright (C) 2018 Linus Walleij + * + * This module binds to the "arm,vexpress-muxfpga" device on the + * Versatile Express configuration bus and sets up which CLCD instance + * gets muxed out on the DVI bridge. + */ +#include <linux/device.h> +#include <linux/module.h> +#include <linux/regmap.h> +#include <linux/vexpress.h> +#include <linux/platform_device.h> +#include <linux/of.h> +#include <linux/of_address.h> +#include <linux/of_platform.h> +#include "pl111_drm.h" +#include "pl111_vexpress.h" + +#define VEXPRESS_FPGAMUX_MOTHERBOARD 0x00 +#define VEXPRESS_FPGAMUX_DAUGHTERBOARD_1 0x01 +#define VEXPRESS_FPGAMUX_DAUGHTERBOARD_2 0x02 + +int pl111_vexpress_clcd_init(struct device *dev, + struct pl111_drm_dev_private *priv, + struct regmap *map) +{ + struct device_node *root; + struct device_node *child; + struct device_node *ct_clcd = NULL; + bool has_coretile_clcd = false; + bool has_coretile_hdlcd = false; + bool mux_motherboard = true; + u32 val; + int ret; + + /* + * Check if we have a CLCD or HDLCD on the core tile by checking if a + * CLCD or HDLCD is available in the root of the device tree. + */ + root = of_find_node_by_path("/"); + if (!root) + return -EINVAL; + + for_each_available_child_of_node(root, child) { + if (of_device_is_compatible(child, "arm,pl111")) { + has_coretile_clcd = true; + ct_clcd = child; + break; + } + if (of_device_is_compatible(child, "arm,hdlcd")) { + has_coretile_hdlcd = true; + break; + } + } + + /* + * If there is a coretile HDLCD and it has a driver, + * do not mux the CLCD on the motherboard to the DVI. + */ + if (has_coretile_hdlcd && IS_ENABLED(CONFIG_DRM_HDLCD)) + mux_motherboard = false; + + /* + * On the Vexpress CA9 we let the CLCD on the coretile + * take precedence, so also in this case do not mux the + * motherboard to the DVI. + */ + if (has_coretile_clcd) + mux_motherboard = false; + + if (mux_motherboard) { + dev_info(dev, "DVI muxed to motherboard CLCD\n"); + val = VEXPRESS_FPGAMUX_MOTHERBOARD; + } else if (ct_clcd == dev->of_node) { + dev_info(dev, + "DVI muxed to daughterboard 1 (core tile) CLCD\n"); + val = VEXPRESS_FPGAMUX_DAUGHTERBOARD_1; + } else { + dev_info(dev, "core tile graphics present\n"); + dev_info(dev, "this device will be deactivated\n"); + return -ENODEV; + } + + ret = regmap_write(map, 0, val); + if (ret) { + dev_err(dev, "error setting DVI muxmode\n"); + return -ENODEV; + } + + return 0; +} + +/* + * This sets up the regmap pointer that will then be retrieved by + * the detection code in pl111_versatile.c and passed in to the + * pl111_vexpress_clcd_init() function above. + */ +static int vexpress_muxfpga_probe(struct platform_device *pdev) +{ + struct device *dev = &pdev->dev; + struct regmap *map; + + map = devm_regmap_init_vexpress_config(&pdev->dev); + if (IS_ERR(map)) + return PTR_ERR(map); + dev_set_drvdata(dev, map); + + return 0; +} + +static const struct of_device_id vexpress_muxfpga_match[] = { + { .compatible = "arm,vexpress-muxfpga", } +}; + +static struct platform_driver vexpress_muxfpga_driver = { + .driver = { + .name = "vexpress-muxfpga", + .of_match_table = of_match_ptr(vexpress_muxfpga_match), + }, + .probe = vexpress_muxfpga_probe, +}; + +module_platform_driver(vexpress_muxfpga_driver); diff --git a/drivers/gpu/drm/pl111/pl111_vexpress.h b/drivers/gpu/drm/pl111/pl111_vexpress.h new file mode 100644 index 000000000000..49876417f7b6 --- /dev/null +++ b/drivers/gpu/drm/pl111/pl111_vexpress.h @@ -0,0 +1,22 @@ +// SPDX-License-Identifier: GPL-2.0 + +struct device; +struct pl111_drm_dev_private; +struct regmap; + +#ifdef CONFIG_ARCH_VEXPRESS + +int pl111_vexpress_clcd_init(struct device *dev, + struct pl111_drm_dev_private *priv, + struct regmap *map); + +#else + +static int inline pl111_vexpress_clcd_init(struct device *dev, + struct pl111_drm_dev_private *priv, + struct regmap *map) +{ + return -ENODEV; +} + +#endif
Hi Linus,
On 23/04/18 08:20, Linus Walleij wrote:
The Versatile Express uses a special configuration controller deeply embedded in the system motherboard FPGA to multiplex the two to three (!) display controller instances out to the single SiI9022 bridge.
Set up an extra file with the logic to probe to the FPGA mux register on the system controller bus, then parse the device tree to see if there is a CLCD or HDLCD instance on the core tile (also known as the daughterboard) by looking in the root of the device tree for compatible nodes.
If there is a HDLCD on the core tile, and there is a driver for it, we exit probe and deactivate the motherboard CLCD. We do not touch the DVI mux in this case, to make sure we don't break HDLCD.
If there is a CLCD on both the motherboard and the core tile (only the CA9 has this) the core tile CLCD takes precedence and get muxed to the DVI connector.
Only if there is no working graphics on the core tile, the motherboard CLCD is probed and muxed to the DVI connector.
Core tile graphics should always take precedence as it can address all memory and is also faster, however the motherboard CLCD is good to have around for diagnostics and testing.
It is possible to test the motherboard CLCD by setting the status = "disabled" property on the core tile CLCD or HDLCD.
Scale down the Versatile Express to 16BPP so we can support a 1024x768 display despite the bus bandwidth restrictions on this platform. (The motherboard CLCD supports slightly lower resolution.)
I've given this a go on the nearest VExpress (with CA15_A7 tile) and sadly it doesn't quite work for me :( With the HDLCD driver configure out, the PL111 driver seems to repeatedly defer somewhere past pl111_versatile_init():
root@jet-set-radio:~# dmesg | grep -i lcd [ 2.739358] drm-clcd-pl111 1c1f0000.clcd: sysreg has not yet probed [ 3.406534] drm-clcd-pl111 1c1f0000.clcd: initializing Versatile Express PL111 [ 3.428218] drm-clcd-pl111 1c1f0000.clcd: DVI muxed to motherboard CLCD [ 3.898768] drm-clcd-pl111 1c1f0000.clcd: initializing Versatile Express PL111 [ 3.920441] drm-clcd-pl111 1c1f0000.clcd: DVI muxed to motherboard CLCD [ 10.896158] drm-clcd-pl111 1c1f0000.clcd: initializing Versatile Express PL111 [ 10.917839] drm-clcd-pl111 1c1f0000.clcd: DVI muxed to motherboard CLCD [ 18.513233] drm-clcd-pl111 1c1f0000.clcd: initializing Versatile Express PL111 [ 18.535007] drm-clcd-pl111 1c1f0000.clcd: DVI muxed to motherboard CLCD root@jet-set-radio:~# ls /sys/bus/amba/drivers/drm-clcd-pl111/ bind uevent unbind
There's no obvious indication of what's missing, but it does still manage to do enough to make the external display come on (still showing the motherboard PoST screen). By comparison, the old fbdev driver probes OK and thinks it's set up a 640x480 console but doesn't actually turn the monitor on (and HDLCD doesn't probe at all due to vexpress-v2p-ca15_a7.dts lacking any graph endpoint).
Config attached, just in case I'm missing something trivially dumb.
Robin.
Cc: Liviu Dudau liviu.dudau@arm.com Cc: Pawel Moll pawel.moll@arm.com Cc: Robin Murphy robin.murphy@arm.com Acked-by: Eric Anholt eric@anholt.net Signed-off-by: Linus Walleij linus.walleij@linaro.org
ChangeLog v3->v4:
- Add a missing platform_device_put()
- Simplify code with for_each_available_child_of_node()
- Collect Eric's ACK.
ChangeLog v2->v3:
- Rewrite CLCD detection and mux priority logic, look in the device tree root for core tile graphics.
ChangeLog v1->v2:
- No changes just reposting rebased on mainline changes.
drivers/gpu/drm/pl111/Makefile | 1 + drivers/gpu/drm/pl111/pl111_versatile.c | 49 ++++++++++++- drivers/gpu/drm/pl111/pl111_vexpress.c | 125 ++++++++++++++++++++++++++++++++ drivers/gpu/drm/pl111/pl111_vexpress.h | 22 ++++++ 4 files changed, 196 insertions(+), 1 deletion(-) create mode 100644 drivers/gpu/drm/pl111/pl111_vexpress.c create mode 100644 drivers/gpu/drm/pl111/pl111_vexpress.h
diff --git a/drivers/gpu/drm/pl111/Makefile b/drivers/gpu/drm/pl111/Makefile index 9c5e8dba8ac6..19a8189dc54f 100644 --- a/drivers/gpu/drm/pl111/Makefile +++ b/drivers/gpu/drm/pl111/Makefile @@ -3,6 +3,7 @@ pl111_drm-y += pl111_display.o \ pl111_versatile.o \ pl111_drv.o
+pl111_drm-$(CONFIG_ARCH_VEXPRESS) += pl111_vexpress.o pl111_drm-$(CONFIG_DEBUG_FS) += pl111_debugfs.o
obj-$(CONFIG_DRM_PL111) += pl111_drm.o diff --git a/drivers/gpu/drm/pl111/pl111_versatile.c b/drivers/gpu/drm/pl111/pl111_versatile.c index 9302f516045e..78ddf8534fd2 100644 --- a/drivers/gpu/drm/pl111/pl111_versatile.c +++ b/drivers/gpu/drm/pl111/pl111_versatile.c @@ -1,12 +1,14 @@ #include <linux/amba/clcd-regs.h> #include <linux/device.h> #include <linux/of.h> +#include <linux/of_platform.h> #include <linux/regmap.h> #include <linux/mfd/syscon.h> #include <linux/bitops.h> #include <linux/module.h> #include <drm/drmP.h> #include "pl111_versatile.h" +#include "pl111_vexpress.h" #include "pl111_drm.h"
static struct regmap *versatile_syscon_map; @@ -22,6 +24,7 @@ enum versatile_clcd { REALVIEW_CLCD_PB11MP, REALVIEW_CLCD_PBA8, REALVIEW_CLCD_PBX,
VEXPRESS_CLCD_V2M, };
static const struct of_device_id versatile_clcd_of_match[] = {
@@ -53,6 +56,10 @@ static const struct of_device_id versatile_clcd_of_match[] = { .compatible = "arm,realview-pbx-syscon", .data = (void *)REALVIEW_CLCD_PBX, },
- {
.compatible = "arm,vexpress-muxfpga",.data = (void *)VEXPRESS_CLCD_V2M,- }, {}, };
@@ -286,12 +293,26 @@ static const struct pl111_variant_data pl111_realview = { .fb_bpp = 16, };
+/*
- Versatile Express PL111 variant, again we just push the maximum
- BPP to 16 to be able to get 1024x768 without saturating the memory
- bus. The clockdivider also seems broken on the Versatile Express.
- */
+static const struct pl111_variant_data pl111_vexpress = {
- .name = "PL111 Versatile Express",
- .formats = pl111_realview_pixel_formats,
- .nformats = ARRAY_SIZE(pl111_realview_pixel_formats),
- .fb_bpp = 16,
- .broken_clockdivider = true,
+};
int pl111_versatile_init(struct device *dev, struct pl111_drm_dev_private *priv) { const struct of_device_id *clcd_id; enum versatile_clcd versatile_clcd_type; struct device_node *np; struct regmap *map;
int ret;
np = of_find_matching_node_and_match(NULL, versatile_clcd_of_match, &clcd_id);
@@ -301,7 +322,26 @@ int pl111_versatile_init(struct device *dev, struct pl111_drm_dev_private *priv) } versatile_clcd_type = (enum versatile_clcd)clcd_id->data;
- map = syscon_node_to_regmap(np);
- /* Versatile Express special handling */
- if (versatile_clcd_type == VEXPRESS_CLCD_V2M) {
struct platform_device *pdev;/* Call into deep Vexpress configuration API */pdev = of_find_device_by_node(np);if (!pdev) {dev_err(dev, "can't find the sysreg device, deferring\n");return -EPROBE_DEFER;}map = dev_get_drvdata(&pdev->dev);if (!map) {dev_err(dev, "sysreg has not yet probed\n");platform_device_put(pdev);return -EPROBE_DEFER;}- } else {
map = syscon_node_to_regmap(np);- }
- if (IS_ERR(map)) { dev_err(dev, "no Versatile syscon regmap\n"); return PTR_ERR(map);
@@ -340,6 +380,13 @@ int pl111_versatile_init(struct device *dev, struct pl111_drm_dev_private *priv) priv->variant_display_disable = pl111_realview_clcd_disable; dev_info(dev, "set up callbacks for RealView PL111\n"); break;
- case VEXPRESS_CLCD_V2M:
priv->variant = &pl111_vexpress;dev_info(dev, "initializing Versatile Express PL111\n");ret = pl111_vexpress_clcd_init(dev, priv, map);if (ret)return ret;break;diff --git a/drivers/gpu/drm/pl111/pl111_vexpress.c b/drivers/gpu/drm/pl111/pl111_vexpress.c new file mode 100644 index 000000000000..9cdb28621ff4 --- /dev/null +++ b/drivers/gpu/drm/pl111/pl111_vexpress.c @@ -0,0 +1,125 @@ +// SPDX-License-Identifier: GPL-2.0 +/*
- Versatile Express PL111 handling
- Copyright (C) 2018 Linus Walleij
- This module binds to the "arm,vexpress-muxfpga" device on the
- Versatile Express configuration bus and sets up which CLCD instance
- gets muxed out on the DVI bridge.
- */
+#include <linux/device.h> +#include <linux/module.h> +#include <linux/regmap.h> +#include <linux/vexpress.h> +#include <linux/platform_device.h> +#include <linux/of.h> +#include <linux/of_address.h> +#include <linux/of_platform.h> +#include "pl111_drm.h" +#include "pl111_vexpress.h"
+#define VEXPRESS_FPGAMUX_MOTHERBOARD 0x00 +#define VEXPRESS_FPGAMUX_DAUGHTERBOARD_1 0x01 +#define VEXPRESS_FPGAMUX_DAUGHTERBOARD_2 0x02
+int pl111_vexpress_clcd_init(struct device *dev,
struct pl111_drm_dev_private *priv,struct regmap *map)+{
- struct device_node *root;
- struct device_node *child;
- struct device_node *ct_clcd = NULL;
- bool has_coretile_clcd = false;
- bool has_coretile_hdlcd = false;
- bool mux_motherboard = true;
- u32 val;
- int ret;
- /*
* Check if we have a CLCD or HDLCD on the core tile by checking if a* CLCD or HDLCD is available in the root of the device tree.*/- root = of_find_node_by_path("/");
- if (!root)
return -EINVAL;- for_each_available_child_of_node(root, child) {
if (of_device_is_compatible(child, "arm,pl111")) {has_coretile_clcd = true;ct_clcd = child;break;}if (of_device_is_compatible(child, "arm,hdlcd")) {has_coretile_hdlcd = true;break;}- }
- /*
* If there is a coretile HDLCD and it has a driver,* do not mux the CLCD on the motherboard to the DVI.*/- if (has_coretile_hdlcd && IS_ENABLED(CONFIG_DRM_HDLCD))
mux_motherboard = false;- /*
* On the Vexpress CA9 we let the CLCD on the coretile* take precedence, so also in this case do not mux the* motherboard to the DVI.*/- if (has_coretile_clcd)
mux_motherboard = false;- if (mux_motherboard) {
dev_info(dev, "DVI muxed to motherboard CLCD\n");val = VEXPRESS_FPGAMUX_MOTHERBOARD;- } else if (ct_clcd == dev->of_node) {
dev_info(dev,"DVI muxed to daughterboard 1 (core tile) CLCD\n");val = VEXPRESS_FPGAMUX_DAUGHTERBOARD_1;- } else {
dev_info(dev, "core tile graphics present\n");dev_info(dev, "this device will be deactivated\n");return -ENODEV;- }
- ret = regmap_write(map, 0, val);
- if (ret) {
dev_err(dev, "error setting DVI muxmode\n");return -ENODEV;- }
- return 0;
+}
+/*
- This sets up the regmap pointer that will then be retrieved by
- the detection code in pl111_versatile.c and passed in to the
- pl111_vexpress_clcd_init() function above.
- */
+static int vexpress_muxfpga_probe(struct platform_device *pdev) +{
- struct device *dev = &pdev->dev;
- struct regmap *map;
- map = devm_regmap_init_vexpress_config(&pdev->dev);
- if (IS_ERR(map))
return PTR_ERR(map);- dev_set_drvdata(dev, map);
- return 0;
+}
+static const struct of_device_id vexpress_muxfpga_match[] = {
- { .compatible = "arm,vexpress-muxfpga", }
+};
+static struct platform_driver vexpress_muxfpga_driver = {
- .driver = {
.name = "vexpress-muxfpga",.of_match_table = of_match_ptr(vexpress_muxfpga_match),- },
- .probe = vexpress_muxfpga_probe,
+};
+module_platform_driver(vexpress_muxfpga_driver); diff --git a/drivers/gpu/drm/pl111/pl111_vexpress.h b/drivers/gpu/drm/pl111/pl111_vexpress.h new file mode 100644 index 000000000000..49876417f7b6 --- /dev/null +++ b/drivers/gpu/drm/pl111/pl111_vexpress.h @@ -0,0 +1,22 @@ +// SPDX-License-Identifier: GPL-2.0
+struct device; +struct pl111_drm_dev_private; +struct regmap;
+#ifdef CONFIG_ARCH_VEXPRESS
+int pl111_vexpress_clcd_init(struct device *dev,
struct pl111_drm_dev_private *priv,struct regmap *map);+#else
+static int inline pl111_vexpress_clcd_init(struct device *dev,
struct pl111_drm_dev_private *priv,struct regmap *map)+{
- return -ENODEV;
+}
+#endif
On Mon, Apr 23, 2018 at 2:04 PM, Robin Murphy robin.murphy@arm.com wrote:
I've given this a go on the nearest VExpress (with CA15_A7 tile) and sadly it doesn't quite work for me :( With the HDLCD driver configure out, the PL111 driver seems to repeatedly defer somewhere past pl111_versatile_init():
root@jet-set-radio:~# dmesg | grep -i lcd [ 2.739358] drm-clcd-pl111 1c1f0000.clcd: sysreg has not yet probed [ 3.406534] drm-clcd-pl111 1c1f0000.clcd: initializing Versatile Express PL111 [ 3.428218] drm-clcd-pl111 1c1f0000.clcd: DVI muxed to motherboard CLCD [ 3.898768] drm-clcd-pl111 1c1f0000.clcd: initializing Versatile Express PL111 [ 3.920441] drm-clcd-pl111 1c1f0000.clcd: DVI muxed to motherboard CLCD [ 10.896158] drm-clcd-pl111 1c1f0000.clcd: initializing Versatile Express PL111 [ 10.917839] drm-clcd-pl111 1c1f0000.clcd: DVI muxed to motherboard CLCD [ 18.513233] drm-clcd-pl111 1c1f0000.clcd: initializing Versatile Express PL111 [ 18.535007] drm-clcd-pl111 1c1f0000.clcd: DVI muxed to motherboard CLCD root@jet-set-radio:~# ls /sys/bus/amba/drivers/drm-clcd-pl111/ bind uevent unbind
There's no obvious indication of what's missing, but it does still manage to do enough to make the external display come on (still showing the motherboard PoST screen). By comparison, the old fbdev driver probes OK and thinks it's set up a 640x480 console but doesn't actually turn the monitor on (and HDLCD doesn't probe at all due to vexpress-v2p-ca15_a7.dts lacking any graph endpoint).
Config attached, just in case I'm missing something trivially dumb.
Hm you definately need CONFIG_CMA and CONFIG_DMA_CMA so the driver actually uses the special memory.
I have all these:
config-drm: config-base $(CURDIR)/scripts/config --file $(config_file) \ --disable FB_ARMCLCD \ --enable CMA \ --enable CMA_DEBUG \ --enable CMA_DEBUGFS \ --enable DMA_CMA \ --enable DRM \ --enable DRM_PL111 \ --enable FRAMEBUFFER_CONSOLE \ --enable DRM_PANEL \ --enable DRM_PANEL_SIMPLE \ --enable DRM_BRIDGE \ --enable DRM_PANEL_BRIDGE \ --enable DRM_SII902X \ --enable LOGO
Please try with the stuff above, when I switch the mainline platforms I plan to update the defconfigs with all the above.
Yours, Linus Walleij
On 26/04/18 08:52, Linus Walleij wrote:
On Mon, Apr 23, 2018 at 2:04 PM, Robin Murphy robin.murphy@arm.com wrote:
I've given this a go on the nearest VExpress (with CA15_A7 tile) and sadly it doesn't quite work for me :( With the HDLCD driver configure out, the PL111 driver seems to repeatedly defer somewhere past pl111_versatile_init():
root@jet-set-radio:~# dmesg | grep -i lcd [ 2.739358] drm-clcd-pl111 1c1f0000.clcd: sysreg has not yet probed [ 3.406534] drm-clcd-pl111 1c1f0000.clcd: initializing Versatile Express PL111 [ 3.428218] drm-clcd-pl111 1c1f0000.clcd: DVI muxed to motherboard CLCD [ 3.898768] drm-clcd-pl111 1c1f0000.clcd: initializing Versatile Express PL111 [ 3.920441] drm-clcd-pl111 1c1f0000.clcd: DVI muxed to motherboard CLCD [ 10.896158] drm-clcd-pl111 1c1f0000.clcd: initializing Versatile Express PL111 [ 10.917839] drm-clcd-pl111 1c1f0000.clcd: DVI muxed to motherboard CLCD [ 18.513233] drm-clcd-pl111 1c1f0000.clcd: initializing Versatile Express PL111 [ 18.535007] drm-clcd-pl111 1c1f0000.clcd: DVI muxed to motherboard CLCD root@jet-set-radio:~# ls /sys/bus/amba/drivers/drm-clcd-pl111/ bind uevent unbind
There's no obvious indication of what's missing, but it does still manage to do enough to make the external display come on (still showing the motherboard PoST screen). By comparison, the old fbdev driver probes OK and thinks it's set up a 640x480 console but doesn't actually turn the monitor on (and HDLCD doesn't probe at all due to vexpress-v2p-ca15_a7.dts lacking any graph endpoint).
Config attached, just in case I'm missing something trivially dumb.
Hm you definately need CONFIG_CMA and CONFIG_DMA_CMA so the driver actually uses the special memory.
Argh, missing CMA was exactly the kind of idiocy I predicted...
I have all these:
config-drm: config-base $(CURDIR)/scripts/config --file $(config_file) \ --disable FB_ARMCLCD \ --enable CMA \ --enable CMA_DEBUG \ --enable CMA_DEBUGFS \ --enable DMA_CMA \ --enable DRM \ --enable DRM_PL111 \ --enable FRAMEBUFFER_CONSOLE \ --enable DRM_PANEL \ --enable DRM_PANEL_SIMPLE \ --enable DRM_BRIDGE \ --enable DRM_PANEL_BRIDGE \ --enable DRM_SII902X \ --enable LOGO
Please try with the stuff above, when I switch the mainline platforms I plan to update the defconfigs with all the above.
...however even ensuring all of the above options I still see the same behaviour, i.e. the driver fails to bind with no visible reason why.
I dug a little bit and it seems pl111_modeset_init() is deferring because it can't find endpoint 0. And given that that's apparently pointing at some non-existent panel rather than the DVI encoder as I would (naively) expect, I can at least form a self-consistent explanation and give up on the grounds that I'm probably missing some important DT changes.
Robin.
On Fri, Apr 27, 2018 at 5:02 PM, Robin Murphy robin.murphy@arm.com wrote:
I dug a little bit and it seems pl111_modeset_init() is deferring because it can't find endpoint 0. And given that that's apparently pointing at some non-existent panel rather than the DVI encoder as I would (naively) expect, I can at least form a self-consistent explanation and give up on the grounds that I'm probably missing some important DT changes.
Yeah it does help if I also send the required DT changes :/
Sorry for my absentmindedness, I'll shoot it off right now so you can test it.
Yours, Linus Walleij
On 27/04/18 19:51, Linus Walleij wrote:
On Fri, Apr 27, 2018 at 5:02 PM, Robin Murphy robin.murphy@arm.com wrote:
I dug a little bit and it seems pl111_modeset_init() is deferring because it can't find endpoint 0. And given that that's apparently pointing at some non-existent panel rather than the DVI encoder as I would (naively) expect, I can at least form a self-consistent explanation and give up on the grounds that I'm probably missing some important DT changes.
Yeah it does help if I also send the required DT changes :/
Sorry for my absentmindedness, I'll shoot it off right now so you can test it.
Hooray, thanks for that.
For these 2 patches with a V2P-CA15_A7 tile and the DT fixed up,
Tested-by: Robin Murphy robin.murphy@arm.com
Framebuffer console and /dev/fb0 access for the motherboard CLCD works, and it keeps out of the way as expected with CONFIG_DRM_HDLCD=y. I can't convince an X session to start, but that appears more to do with sii902x falling apart*, which I can't be bothered to even try debugging.
Robin.
*specifically, lots of this, which I imagine is related to xorg trying to either read the EDID or change mode: ... [ 243.439669] i2c i2c-0: sendbytes: NAK bailout. [ 243.454247] i2c i2c-0: sendbytes: NAK bailout. [ 247.743261] i2c i2c-0: sendbytes: NAK bailout. [ 247.770281] sii902x 0-0039: failed to read status (-6) [ 247.786609] i2c i2c-0: sendbytes: NAK bailout. [ 248.118522] i2c i2c-0: sendbytes: NAK bailout. [ 248.145485] sii902x 0-0039: failed to read status (-6) [ 248.177800] i2c i2c-0: sendbytes: NAK bailout. [ 258.256500] i2c i2c-0: sendbytes: NAK bailout. [ 258.274763] i2c i2c-0: sendbytes: NAK bailout. [ 268.499637] i2c i2c-0: sendbytes: NAK bailout. [ 268.517899] i2c i2c-0: sendbytes: NAK bailout. ...
On Mon, Apr 30, 2018 at 03:55:41PM +0100, Robin Murphy wrote:
On 27/04/18 19:51, Linus Walleij wrote:
On Fri, Apr 27, 2018 at 5:02 PM, Robin Murphy robin.murphy@arm.com wrote:
I dug a little bit and it seems pl111_modeset_init() is deferring because it can't find endpoint 0. And given that that's apparently pointing at some non-existent panel rather than the DVI encoder as I would (naively) expect, I can at least form a self-consistent explanation and give up on the grounds that I'm probably missing some important DT changes.
Yeah it does help if I also send the required DT changes :/
Sorry for my absentmindedness, I'll shoot it off right now so you can test it.
Hooray, thanks for that.
For these 2 patches with a V2P-CA15_A7 tile and the DT fixed up,
Tested-by: Robin Murphy robin.murphy@arm.com
Framebuffer console and /dev/fb0 access for the motherboard CLCD works, and it keeps out of the way as expected with CONFIG_DRM_HDLCD=y. I can't convince an X session to start, but that appears more to do with sii902x falling apart*, which I can't be bothered to even try debugging.
Robin.
*specifically, lots of this, which I imagine is related to xorg trying to either read the EDID or change mode: ... [ 243.439669] i2c i2c-0: sendbytes: NAK bailout. [ 243.454247] i2c i2c-0: sendbytes: NAK bailout. [ 247.743261] i2c i2c-0: sendbytes: NAK bailout. [ 247.770281] sii902x 0-0039: failed to read status (-6) [ 247.786609] i2c i2c-0: sendbytes: NAK bailout. [ 248.118522] i2c i2c-0: sendbytes: NAK bailout. [ 248.145485] sii902x 0-0039: failed to read status (-6) [ 248.177800] i2c i2c-0: sendbytes: NAK bailout. [ 258.256500] i2c i2c-0: sendbytes: NAK bailout. [ 258.274763] i2c i2c-0: sendbytes: NAK bailout. [ 268.499637] i2c i2c-0: sendbytes: NAK bailout. [ 268.517899] i2c i2c-0: sendbytes: NAK bailout.
That's a long standing issue that I've never got to the bottom of. The sii902x seems to have been wired in a curious way on the VExpress board and it is struggling to pass-through the I2C connection to the monitor.
Best regards, Liviu
... _______________________________________________ dri-devel mailing list dri-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/dri-devel
dri-devel@lists.freedesktop.org
-
Linus Walleij -
Liviu Dudau -
Robin Murphy