mail archive of the barebox mailing list
 help / color / mirror / Atom feed
* [RESEND PATCH] framebuffer work (especially i.MX)
@ 2013-01-17  6:32 Daniel Mierswa
  2013-01-17  6:32 ` [PATCH 1/4] i.MX21: Add periph. clock register name macros Daniel Mierswa
                   ` (4 more replies)
  0 siblings, 5 replies; 10+ messages in thread
From: Daniel Mierswa @ 2013-01-17  6:32 UTC (permalink / raw)
  To: barebox

I adjusted the patch series and added the fixes as discussed on the
mailinglist.

I still somehow feel that the offscreenbuf member should be set by
the fd_open routine because I expect fd_close to always work with the
struct set by fd_open no matter what offscreen parameter the function
got.

Daniel Mierswa (4):
  i.MX21: Add periph. clock register name macros
  i.MX21/27: don't enable lcd bus clocks too early
  i.MX27: fix shift amount for PCCR1_PERCLK3_EN
  video/imx: initialize offscreenbuf member

 arch/arm/mach-imx/clk-imx21.c               | 67 ++++++++++++++++++++++++++---
 arch/arm/mach-imx/clk-imx27.c               | 16 ++++---
 arch/arm/mach-imx/include/mach/imx21-regs.h |  6 ---
 commands/splash.c                           |  8 ++++
 drivers/video/imx.c                         | 40 ++++++++++++++---
 5 files changed, 111 insertions(+), 26 deletions(-)

-- 
1.8.1


_______________________________________________
barebox mailing list
barebox@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/barebox

^ permalink raw reply	[flat|nested] 10+ messages in thread

* [PATCH 1/4] i.MX21: Add periph. clock register name macros
  2013-01-17  6:32 [RESEND PATCH] framebuffer work (especially i.MX) Daniel Mierswa
@ 2013-01-17  6:32 ` Daniel Mierswa
  2013-01-17  6:32 ` [PATCH 2/4] i.MX21/27: don't enable lcd bus clocks too early Daniel Mierswa
                   ` (3 subsequent siblings)
  4 siblings, 0 replies; 10+ messages in thread
From: Daniel Mierswa @ 2013-01-17  6:32 UTC (permalink / raw)
  To: barebox

Also put those names solely in the .c file as it's done with
the i.MX27 code.

Signed-off-by: Daniel Mierswa <d.mierswa@phytec.de>
---
 arch/arm/mach-imx/clk-imx21.c               | 56 +++++++++++++++++++++++++----
 arch/arm/mach-imx/include/mach/imx21-regs.h |  6 ----
 2 files changed, 50 insertions(+), 12 deletions(-)

diff --git a/arch/arm/mach-imx/clk-imx21.c b/arch/arm/mach-imx/clk-imx21.c
index 9e7af81..784951d 100644
--- a/arch/arm/mach-imx/clk-imx21.c
+++ b/arch/arm/mach-imx/clk-imx21.c
@@ -45,6 +45,48 @@
 #define CCM_PMCOUNT		0x30
 #define CCM_WKGDCTL		0x34
 
+#define PCCR0_UART1_EN		(1 << 0)
+#define PCCR0_UART2_EN		(1 << 1)
+#define PCCR0_UART3_EN		(1 << 2)
+#define PCCR0_UART4_EN		(1 << 3)
+#define PCCR0_CSPI1_EN		(1 << 4)
+#define PCCR0_CSPI2_EN		(1 << 5)
+#define PCCR0_SSI1_EN		(1 << 6)
+#define PCCR0_SSI2_EN		(1 << 7)
+#define PCCR0_FIRI_EN		(1 << 8)
+#define PCCR0_SDHC1_EN		(1 << 9)
+#define PCCR0_SDHC2_EN		(1 << 10)
+#define PCCR0_GPIO_EN		(1 << 11)
+#define PCCR0_I2C_EN		(1 << 12)
+#define PCCR0_DMA_EN		(1 << 13)
+#define PCCR0_USBOTG_EN		(1 << 14)
+#define PCCR0_EMMA_EN		(1 << 15)
+#define PCCR0_SSI2_BAUD_EN	(1 << 16)
+#define PCCR0_SSI1_BAUD_EN	(1 << 17)
+#define PCCR0_PERCLK3_EN	(1 << 18)
+#define PCCR0_NFC_EN		(1 << 19)
+#define PCCR0_FRI_BAUD_EN	(1 << 20)
+#define PCCR0_SLDC_EN		(1 << 21)
+#define PCCR0_PERCLK4_EN	(1 << 22)
+#define PCCR0_HCLK_BMI_EN	(1 << 23)
+#define PCCR0_HCLK_USBOTG_EN	(1 << 24)
+#define PCCR0_HCLK_SLCDC_EN	(1 << 25)
+#define PCCR0_HCLK_LCDC_EN	(1 << 26)
+#define PCCR0_HCLK_EMMA_EN	(1 << 27)
+#define PCCR0_HCLK_BROM_EN	(1 << 28)
+#define PCCR0_HCLK_DMA_EN	(1 << 30)
+#define PCCR0_HCLK_CSI_EN	(1 << 31)
+
+#define PCCR1_CSPI3_EN	(1 << 23)
+#define PCCR1_WDT_EN	(1 << 24)
+#define PCCR1_GPT1_EN	(1 << 25)
+#define PCCR1_GPT2_EN	(1 << 26)
+#define PCCR1_GPT3_EN	(1 << 27)
+#define PCCR1_PWM_EN	(1 << 28)
+#define PCCR1_RTC_EN	(1 << 29)
+#define PCCR1_KPP_EN	(1 << 30)
+#define PCCR1_OWIRE_EN	(1 << 31)
+
 enum imx21_clks {
 	ckil, ckih, fpm, mpll_sel, spll_sel, mpll, spll, fclk, hclk, ipg, per1,
 	per2, per3, per4, usb_div, nfc_div, lcdc_per_gate, clk_max
@@ -70,14 +112,16 @@ static int imx21_ccm_probe(struct device_d *dev)
 
 	base = dev_request_mem_region(dev, 0);
 
-	writel((1 << 0) | (1 << 1) | (1 << 2) | (1 << 3) | (1 << 4) | (1 << 5) |
-			(1 << 9) | (1 << 10) | (1 << 11) | (1 << 12) |
-			(1 << 13) | (1 << 14) | (1 << 19) | (1 << 22) |
-			(1 << 24) | (1 << 26) |	(1 << 30),
+	writel(PCCR0_UART1_EN | PCCR0_UART2_EN | PCCR0_UART3_EN | PCCR0_UART4_EN |
+			PCCR0_CSPI1_EN | PCCR0_CSPI2_EN | PCCR0_SDHC1_EN |
+			PCCR0_SDHC2_EN | PCCR0_GPIO_EN | PCCR0_I2C_EN | PCCR0_DMA_EN |
+			PCCR0_USBOTG_EN | PCCR0_NFC_EN | PCCR0_PERCLK4_EN |
+			PCCR0_HCLK_USBOTG_EN | PCCR0_HCLK_LCDC_EN | PCCR0_HCLK_DMA_EN,
 			base + CCM_PCCR0);
 
-	writel((1 << 23) | (1 << 24) | (1 << 25) | (1 << 26) | (1 << 27) |
-			(1 << 28) | (1 << 29) | (1 << 30) | (1 << 31),
+	writel(PCCR1_CSPI3_EN | PCCR1_WDT_EN | PCCR1_GPT1_EN | PCCR1_GPT2_EN |
+			PCCR1_GPT3_EN | PCCR1_PWM_EN | PCCR1_RTC_EN | PCCR1_KPP_EN |
+			PCCR1_OWIRE_EN,
 			base + CCM_PCCR1);
 
 	clks[ckil] = clk_fixed("ckil", lref);
diff --git a/arch/arm/mach-imx/include/mach/imx21-regs.h b/arch/arm/mach-imx/include/mach/imx21-regs.h
index 1c4b550..87bd99c 100644
--- a/arch/arm/mach-imx/include/mach/imx21-regs.h
+++ b/arch/arm/mach-imx/include/mach/imx21-regs.h
@@ -129,12 +129,6 @@
 #define MX21_MPCTL1_BRMO	(1 << 6)
 #define MX21_MPCTL1_LF		(1 << 15)
 
-#define MX21_PCCR0_PERCLK3_EN	(1 << 18)
-#define MX21_PCCR0_NFC_EN	(1 << 19)
-#define MX21_PCCR0_HCLK_LCDC_EN	(1 << 26)
-
-#define MX21_PCCR1_GPT1_EN	(1 << 25)
-
 #define MX21_CCSR_32K_SR	(1 << 15)
 
 #endif /* _IMX21_REGS_H */
-- 
1.8.1


_______________________________________________
barebox mailing list
barebox@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/barebox

^ permalink raw reply	[flat|nested] 10+ messages in thread

* [PATCH 2/4] i.MX21/27: don't enable lcd bus clocks too early
  2013-01-17  6:32 [RESEND PATCH] framebuffer work (especially i.MX) Daniel Mierswa
  2013-01-17  6:32 ` [PATCH 1/4] i.MX21: Add periph. clock register name macros Daniel Mierswa
@ 2013-01-17  6:32 ` Daniel Mierswa
  2013-01-17  6:32 ` [PATCH 3/4] i.MX27: fix shift amount for PCCR1_PERCLK3_EN Daniel Mierswa
                   ` (2 subsequent siblings)
  4 siblings, 0 replies; 10+ messages in thread
From: Daniel Mierswa @ 2013-01-17  6:32 UTC (permalink / raw)
  To: barebox

On the MX27 based board phycard-i.MX27 the display won't properly
come up.
Before removing imx-regs.h and the code that sets the register
in the i.MX video driver, the PCCR registers were set _after_
the screen start (LSSAR) was set.
This restores that old behaviour and makes the display come up
properly again.
I did not have a chance to test this on any other i.MX27 or i.MX21
hardware though I assume that the "old" order is required there
too.

Signed-off-by: Daniel Mierswa <d.mierswa@phytec.de>
---
 arch/arm/mach-imx/clk-imx21.c | 13 +++++++++++--
 arch/arm/mach-imx/clk-imx27.c | 14 +++++++++-----
 drivers/video/imx.c           | 40 +++++++++++++++++++++++++++++++++-------
 3 files changed, 53 insertions(+), 14 deletions(-)

diff --git a/arch/arm/mach-imx/clk-imx21.c b/arch/arm/mach-imx/clk-imx21.c
index 784951d..6e91424 100644
--- a/arch/arm/mach-imx/clk-imx21.c
+++ b/arch/arm/mach-imx/clk-imx21.c
@@ -89,7 +89,8 @@
 
 enum imx21_clks {
 	ckil, ckih, fpm, mpll_sel, spll_sel, mpll, spll, fclk, hclk, ipg, per1,
-	per2, per3, per4, usb_div, nfc_div, lcdc_per_gate, clk_max
+	per2, per3, per4, usb_div, nfc_div, lcdc_per_gate, lcdc_ahb_gate,
+	lcdc_ipg_gate, clk_max
 };
 
 static struct clk *clks[clk_max];
@@ -116,7 +117,7 @@ static int imx21_ccm_probe(struct device_d *dev)
 			PCCR0_CSPI1_EN | PCCR0_CSPI2_EN | PCCR0_SDHC1_EN |
 			PCCR0_SDHC2_EN | PCCR0_GPIO_EN | PCCR0_I2C_EN | PCCR0_DMA_EN |
 			PCCR0_USBOTG_EN | PCCR0_NFC_EN | PCCR0_PERCLK4_EN |
-			PCCR0_HCLK_USBOTG_EN | PCCR0_HCLK_LCDC_EN | PCCR0_HCLK_DMA_EN,
+			PCCR0_HCLK_USBOTG_EN | PCCR0_HCLK_DMA_EN,
 			base + CCM_PCCR0);
 
 	writel(PCCR1_CSPI3_EN | PCCR1_WDT_EN | PCCR1_GPT1_EN | PCCR1_GPT2_EN |
@@ -143,6 +144,12 @@ static int imx21_ccm_probe(struct device_d *dev)
 	clks[usb_div] = imx_clk_divider("usb_div", "spll", base + CCM_CSCR, 26, 3);
 	clks[nfc_div] = imx_clk_divider("nfc_div", "ipg", base + CCM_PCDR0, 12, 4);
 	clks[lcdc_per_gate] = imx_clk_gate("lcdc_per_gate", "per3", base + CCM_PCCR0, 18);
+	clks[lcdc_ahb_gate] = imx_clk_gate("lcdc_ahb_gate", "ahb", base + CCM_PCCR0, 26);
+	/*
+	 * i.MX21 doesn't have an IPG clock for the LCD. To avoid even more conditionals
+	 * in the framebuffer code, provide a dummy clock.
+	 */
+	clks[lcdc_ipg_gate] = clk_fixed("dummy", 0);
 
 	clkdev_add_physbase(clks[per1], MX21_GPT1_BASE_ADDR, NULL);
 	clkdev_add_physbase(clks[per1], MX21_GPT2_BASE_ADDR, NULL);
@@ -158,6 +165,8 @@ static int imx21_ccm_probe(struct device_d *dev)
 	clkdev_add_physbase(clks[ipg], MX21_SDHC1_BASE_ADDR, NULL);
 	clkdev_add_physbase(clks[ipg], MX21_SDHC2_BASE_ADDR, NULL);
 	clkdev_add_physbase(clks[lcdc_per_gate], MX21_LCDC_BASE_ADDR, NULL);
+	clkdev_add_physbase(clks[lcdc_ahb_gate], MX21_LCDC_BASE_ADDR, "ahb");
+	clkdev_add_physbase(clks[lcdc_ipg_gate], MX21_LCDC_BASE_ADDR, "ipg");
 
 	return 0;
 }
diff --git a/arch/arm/mach-imx/clk-imx27.c b/arch/arm/mach-imx/clk-imx27.c
index 222d2a6..958495e 100644
--- a/arch/arm/mach-imx/clk-imx27.c
+++ b/arch/arm/mach-imx/clk-imx27.c
@@ -93,7 +93,7 @@
 enum mx27_clks {
 	dummy, ckih, ckil, mpll, spll, mpll_main2, ahb, ipg, nfc_div, per1_div,
 	per2_div, per3_div, per4_div, usb_div, cpu_sel,	clko_sel, cpu_div, clko_div,
-	clko_en, lcdc_per_gate, clk_max
+	clko_en, lcdc_per_gate, lcdc_ahb_gate, lcdc_ipg_gate, clk_max
 };
 
 static struct clk *clks[clk_max];
@@ -136,7 +136,7 @@ static int imx27_ccm_probe(struct device_d *dev)
 	base = dev_request_mem_region(dev, 0);
 
 	writel(PCCR0_SDHC3_EN | PCCR0_SDHC2_EN | PCCR0_SDHC1_EN |
-			PCCR0_PWM_EN | PCCR0_KPP_EN | PCCR0_LCDC_EN | PCCR0_IIM_EN |
+			PCCR0_PWM_EN | PCCR0_KPP_EN | PCCR0_IIM_EN |
 			PCCR0_I2C2_EN | PCCR0_I2C1_EN | PCCR0_GPT6_EN | PCCR0_GPT5_EN |
 			PCCR0_GPT4_EN | PCCR0_GPT3_EN | PCCR0_GPT2_EN | PCCR0_GPT1_EN |
 			PCCR0_GPIO_EN | PCCR0_FEC_EN | PCCR0_CSPI3_EN | PCCR0_CSPI2_EN |
@@ -144,9 +144,9 @@ static int imx27_ccm_probe(struct device_d *dev)
 			base + CCM_PCCR0);
 
 	writel(PCCR1_NFC_BAUDEN | PCCR1_PERCLK4_EN | PCCR1_PERCLK2_EN | PCCR1_PERCLK1_EN |
-			PCCR1_HCLK_USB | PCCR1_HCLK_LCDC | PCCR1_HCLK_FEC | PCCR1_HCLK_EMI |
-			PCCR1_WDT_EN | PCCR1_USB_EN | PCCR1_UART6_EN | PCCR1_UART5_EN |
-			PCCR1_UART4_EN | PCCR1_UART3_EN | PCCR1_UART2_EN | PCCR1_UART1_EN,
+			PCCR1_HCLK_USB | PCCR1_HCLK_FEC | PCCR1_HCLK_EMI | PCCR1_WDT_EN |
+			PCCR1_USB_EN | PCCR1_UART6_EN | PCCR1_UART5_EN | PCCR1_UART4_EN |
+			PCCR1_UART3_EN | PCCR1_UART2_EN | PCCR1_UART1_EN,
 			base + CCM_PCCR1);
 
 	clks[dummy] = clk_fixed("dummy", 0);
@@ -180,6 +180,8 @@ static int imx27_ccm_probe(struct device_d *dev)
 		clks[cpu_div] = imx_clk_divider("cpu_div", "cpu_sel", base + CCM_CSCR, 13, 3);
 	clks[clko_div] = imx_clk_divider("clko_div", "clko_sel", base + CCM_PCDR0, 22, 3);
 	clks[lcdc_per_gate] = imx_clk_gate("lcdc_per_gate", "per3_div", base + CCM_PCCR1, 7);
+	clks[lcdc_ahb_gate] = imx_clk_gate("lcdc_ahb_gate", "ahb", base + CCM_PCCR1, 15);
+	clks[lcdc_ipg_gate] = imx_clk_gate("lcdc_ipg_gate", "ipg", base + CCM_PCCR0, 14);
 
 	clkdev_add_physbase(clks[per1_div], MX27_GPT1_BASE_ADDR, NULL);
 	clkdev_add_physbase(clks[per1_div], MX27_GPT2_BASE_ADDR, NULL);
@@ -202,6 +204,8 @@ static int imx27_ccm_probe(struct device_d *dev)
 	clkdev_add_physbase(clks[per2_div], MX27_SDHC2_BASE_ADDR, NULL);
 	clkdev_add_physbase(clks[per2_div], MX27_SDHC3_BASE_ADDR, NULL);
 	clkdev_add_physbase(clks[lcdc_per_gate], MX27_LCDC_BASE_ADDR, NULL);
+	clkdev_add_physbase(clks[lcdc_ahb_gate], MX27_LCDC_BASE_ADDR, "ahb");
+	clkdev_add_physbase(clks[lcdc_ipg_gate], MX27_LCDC_BASE_ADDR, "ipg");
 	clkdev_add_physbase(clks[ipg], MX27_FEC_BASE_ADDR, NULL);
 
 	return 0;
diff --git a/drivers/video/imx.c b/drivers/video/imx.c
index 8381690..54fb700 100644
--- a/drivers/video/imx.c
+++ b/drivers/video/imx.c
@@ -138,7 +138,13 @@ struct imxfb_rgb {
 
 struct imxfb_info {
 	void __iomem		*regs;
-	struct clk		*clk;
+
+#if defined(CONFIG_ARCH_IMX21) || defined(CONFIG_ARCH_IMX27)
+	struct clk		*ahb_clk;
+	struct clk		*ipg_clk;
+#endif
+
+	struct clk		*per_clk;
 
 	u_int			pcr;
 	u_int			pwmr;
@@ -252,7 +258,12 @@ static void imxfb_enable_controller(struct fb_info *info)
 
 	writel(RMCR_LCDC_EN, fbi->regs + LCDC_RMCR);
 
-	clk_enable(fbi->clk);
+#if defined(CONFIG_ARCH_IMX21) || defined(CONFIG_ARCH_IMX27)
+	clk_enable(fbi->ahb_clk);
+	clk_enable(fbi->ipg_clk);
+#endif
+
+	clk_enable(fbi->per_clk);
 
 	if (fbi->enable)
 		fbi->enable(1);
@@ -267,7 +278,12 @@ static void imxfb_disable_controller(struct fb_info *info)
 
 	writel(0, fbi->regs + LCDC_RMCR);
 
-	clk_disable(fbi->clk);
+	clk_disable(fbi->per_clk);
+
+#if defined(CONFIG_ARCH_IMX21) || defined(CONFIG_ARCH_IMX27)
+	clk_disable(fbi->ipg_clk);
+	clk_disable(fbi->ahb_clk);
+#endif
 }
 
 static void set_reg_overlay(struct fb_info *overlay)
@@ -335,7 +351,7 @@ static int imxfb_activate_var(struct fb_info *info)
 	writel(readl(fbi->regs + LCDC_CPOS) & ~(CPOS_CC0 | CPOS_CC1),
 		fbi->regs + LCDC_CPOS);
 
-	lcd_clk = clk_get_rate(fbi->clk);
+	lcd_clk = clk_get_rate(fbi->per_clk);
 
 	tmp = mode->pixclock * (unsigned long long)lcd_clk;
 
@@ -549,9 +565,19 @@ static int imxfb_probe(struct device_d *dev)
 	fbi = xzalloc(sizeof(*fbi));
 	info = &fbi->info;
 
-	fbi->clk = clk_get(dev, NULL);
-	if (IS_ERR(fbi->clk))
-		return PTR_ERR(fbi->clk);
+	fbi->per_clk = clk_get(dev, NULL);
+	if (IS_ERR(fbi->per_clk))
+		return PTR_ERR(fbi->per_clk);
+
+#if defined(CONFIG_ARCH_IMX21) || defined(CONFIG_ARCH_IMX27)
+	fbi->ahb_clk = clk_get(dev, "ahb");
+	if (IS_ERR(fbi->ahb_clk))
+		return PTR_ERR(fbi->ahb_clk);
+
+	fbi->ipg_clk = clk_get(dev, "ipg");
+	if (IS_ERR(fbi->ipg_clk))
+		return PTR_ERR(fbi->ipg_clk);
+#endif
 
 	fbi->mode = pdata->mode;
 	fbi->regs = dev_request_mem_region(dev, 0);
-- 
1.8.1


_______________________________________________
barebox mailing list
barebox@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/barebox

^ permalink raw reply	[flat|nested] 10+ messages in thread

* [PATCH 3/4] i.MX27: fix shift amount for PCCR1_PERCLK3_EN
  2013-01-17  6:32 [RESEND PATCH] framebuffer work (especially i.MX) Daniel Mierswa
  2013-01-17  6:32 ` [PATCH 1/4] i.MX21: Add periph. clock register name macros Daniel Mierswa
  2013-01-17  6:32 ` [PATCH 2/4] i.MX21/27: don't enable lcd bus clocks too early Daniel Mierswa
@ 2013-01-17  6:32 ` Daniel Mierswa
  2013-01-17  6:32 ` [PATCH 4/4] video/imx: always initialize offscreenbuf member Daniel Mierswa
  2013-01-18 10:23 ` [RESEND PATCH] framebuffer work (especially i.MX) Sascha Hauer
  4 siblings, 0 replies; 10+ messages in thread
From: Daniel Mierswa @ 2013-01-17  6:32 UTC (permalink / raw)
  To: barebox


Signed-off-by: Daniel Mierswa <d.mierswa@phytec.de>
---
 arch/arm/mach-imx/clk-imx27.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/arch/arm/mach-imx/clk-imx27.c b/arch/arm/mach-imx/clk-imx27.c
index 958495e..e221928 100644
--- a/arch/arm/mach-imx/clk-imx27.c
+++ b/arch/arm/mach-imx/clk-imx27.c
@@ -179,7 +179,7 @@ static int imx27_ccm_probe(struct device_d *dev)
 	else
 		clks[cpu_div] = imx_clk_divider("cpu_div", "cpu_sel", base + CCM_CSCR, 13, 3);
 	clks[clko_div] = imx_clk_divider("clko_div", "clko_sel", base + CCM_PCDR0, 22, 3);
-	clks[lcdc_per_gate] = imx_clk_gate("lcdc_per_gate", "per3_div", base + CCM_PCCR1, 7);
+	clks[lcdc_per_gate] = imx_clk_gate("lcdc_per_gate", "per3_div", base + CCM_PCCR1, 8);
 	clks[lcdc_ahb_gate] = imx_clk_gate("lcdc_ahb_gate", "ahb", base + CCM_PCCR1, 15);
 	clks[lcdc_ipg_gate] = imx_clk_gate("lcdc_ipg_gate", "ipg", base + CCM_PCCR0, 14);
 
-- 
1.8.1


_______________________________________________
barebox mailing list
barebox@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/barebox

^ permalink raw reply	[flat|nested] 10+ messages in thread

* [PATCH 4/4] video/imx: always initialize offscreenbuf member
  2013-01-17  6:32 [RESEND PATCH] framebuffer work (especially i.MX) Daniel Mierswa
                   ` (2 preceding siblings ...)
  2013-01-17  6:32 ` [PATCH 3/4] i.MX27: fix shift amount for PCCR1_PERCLK3_EN Daniel Mierswa
@ 2013-01-17  6:32 ` Daniel Mierswa
  2013-01-17 11:56   ` Jean-Christophe PLAGNIOL-VILLARD
  2013-01-18 10:23 ` [RESEND PATCH] framebuffer work (especially i.MX) Sascha Hauer
  4 siblings, 1 reply; 10+ messages in thread
From: Daniel Mierswa @ 2013-01-17  6:32 UTC (permalink / raw)
  To: barebox

If offscreen was not passed to fd_open the resulting offscreenbuf
member was a dangling pointer and the free() call in fd_close
would result in undefined behaviour.

Signed-off-by: Daniel Mierswa <d.mierswa@phytec.de>
---
 commands/splash.c | 8 ++++++++
 1 file changed, 8 insertions(+)

diff --git a/commands/splash.c b/commands/splash.c
index 4cc463e..75c7074 100644
--- a/commands/splash.c
+++ b/commands/splash.c
@@ -59,6 +59,14 @@ static int do_splash(int argc, char *argv[])
 		return 1;
 	}
 
+	/*
+	 * sc.offscreenbuf is dangling here, if no offscreen buffer was
+	 * requested, so set it to NULL otherwise the free in fd_close
+	 * will cause undefined behaviour
+	 */
+	if (!offscreen)
+		sc.offscreenbuf = NULL;
+
 	if (sc.offscreenbuf) {
 		if (do_bg)
 			memset_pixel(&info, sc.offscreenbuf, bg_color,
-- 
1.8.1


_______________________________________________
barebox mailing list
barebox@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/barebox

^ permalink raw reply	[flat|nested] 10+ messages in thread

* Re: [PATCH 4/4] video/imx: always initialize offscreenbuf member
  2013-01-17  6:32 ` [PATCH 4/4] video/imx: always initialize offscreenbuf member Daniel Mierswa
@ 2013-01-17 11:56   ` Jean-Christophe PLAGNIOL-VILLARD
  2013-01-17 13:04     ` Daniel Mierswa
  0 siblings, 1 reply; 10+ messages in thread
From: Jean-Christophe PLAGNIOL-VILLARD @ 2013-01-17 11:56 UTC (permalink / raw)
  To: Daniel Mierswa; +Cc: barebox

On 07:32 Thu 17 Jan     , Daniel Mierswa wrote:
> If offscreen was not passed to fd_open the resulting offscreenbuf
> member was a dangling pointer and the free() call in fd_close
> would result in undefined behaviour.
> 
> Signed-off-by: Daniel Mierswa <d.mierswa@phytec.de>
> ---
>  commands/splash.c | 8 ++++++++
>  1 file changed, 8 insertions(+)
> 
> diff --git a/commands/splash.c b/commands/splash.c
> index 4cc463e..75c7074 100644
> --- a/commands/splash.c
> +++ b/commands/splash.c
> @@ -59,6 +59,14 @@ static int do_splash(int argc, char *argv[])
>  		return 1;
>  	}
>  
> +	/*
> +	 * sc.offscreenbuf is dangling here, if no offscreen buffer was
> +	 * requested, so set it to NULL otherwise the free in fd_close
> +	 * will cause undefined behaviour
> +	 */
> +	if (!offscreen)
> +		sc.offscreenbuf = NULL;
> +
NAck

this is already done by memset

Best Regards,
J.
>  	if (sc.offscreenbuf) {
>  		if (do_bg)
>  			memset_pixel(&info, sc.offscreenbuf, bg_color,
> -- 
> 1.8.1
> 
> 
> _______________________________________________
> barebox mailing list
> barebox@lists.infradead.org
> http://lists.infradead.org/mailman/listinfo/barebox

_______________________________________________
barebox mailing list
barebox@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/barebox

^ permalink raw reply	[flat|nested] 10+ messages in thread

* Re: [PATCH 4/4] video/imx: always initialize offscreenbuf member
  2013-01-17 11:56   ` Jean-Christophe PLAGNIOL-VILLARD
@ 2013-01-17 13:04     ` Daniel Mierswa
  2013-01-17 17:55       ` Sascha Hauer
  0 siblings, 1 reply; 10+ messages in thread
From: Daniel Mierswa @ 2013-01-17 13:04 UTC (permalink / raw)
  To: barebox

On 01/17/2013 12:56 PM, Jean-Christophe PLAGNIOL-VILLARD wrote:
> On 07:32 Thu 17 Jan     , Daniel Mierswa wrote:
>> +	/*
>> +	 * sc.offscreenbuf is dangling here, if no offscreen buffer was
>> +	 * requested, so set it to NULL otherwise the free in fd_close
>> +	 * will cause undefined behaviour
>> +	 */
>> +	if (!offscreen)
>> +		sc.offscreenbuf = NULL;
>> +
> NAck
> 
> this is already done by memset
> 
> Best Regards,
> J.
>>  	if (sc.offscreenbuf) {
>>  		if (do_bg)
>>  			memset_pixel(&info, sc.offscreenbuf, bg_color,

Huh? I don't get it.

-- 
Mierswa, Daniel

If you still don't like it, that's ok: that's why I'm boss. I simply know better than you do.
               --- Linus Torvalds, comp.os.linux.advocacy, 1996/07/22

_______________________________________________
barebox mailing list
barebox@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/barebox

^ permalink raw reply	[flat|nested] 10+ messages in thread

* Re: [PATCH 4/4] video/imx: always initialize offscreenbuf member
  2013-01-17 13:04     ` Daniel Mierswa
@ 2013-01-17 17:55       ` Sascha Hauer
  2013-01-18  4:24         ` Daniel Mierswa
  0 siblings, 1 reply; 10+ messages in thread
From: Sascha Hauer @ 2013-01-17 17:55 UTC (permalink / raw)
  To: Daniel Mierswa; +Cc: barebox

On Thu, Jan 17, 2013 at 02:04:22PM +0100, Daniel Mierswa wrote:
> On 01/17/2013 12:56 PM, Jean-Christophe PLAGNIOL-VILLARD wrote:
> > On 07:32 Thu 17 Jan     , Daniel Mierswa wrote:
> >> +	/*
> >> +	 * sc.offscreenbuf is dangling here, if no offscreen buffer was
> >> +	 * requested, so set it to NULL otherwise the free in fd_close
> >> +	 * will cause undefined behaviour
> >> +	 */
> >> +	if (!offscreen)
> >> +		sc.offscreenbuf = NULL;
> >> +
> > NAck
> > 
> > this is already done by memset
> > 
> > Best Regards,
> > J.
> >>  	if (sc.offscreenbuf) {
> >>  		if (do_bg)
> >>  			memset_pixel(&info, sc.offscreenbuf, bg_color,
> 
> Huh? I don't get it.

sc is initialized to 0 in do_splash:

	memset(&s, 0, sizeof(s));
	memset(&sc, 0, sizeof(sc));
	memset(&info, 0, sizeof(info));

so sc.offscreenbuf = NULL is a no-op.

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 |

_______________________________________________
barebox mailing list
barebox@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/barebox

^ permalink raw reply	[flat|nested] 10+ messages in thread

* Re: [PATCH 4/4] video/imx: always initialize offscreenbuf member
  2013-01-17 17:55       ` Sascha Hauer
@ 2013-01-18  4:24         ` Daniel Mierswa
  0 siblings, 0 replies; 10+ messages in thread
From: Daniel Mierswa @ 2013-01-18  4:24 UTC (permalink / raw)
  To: barebox

On 01/17/2013 06:55 PM, Sascha Hauer wrote:
> sc is initialized to 0 in do_splash:
> 
> 	memset(&s, 0, sizeof(s));
> 	memset(&sc, 0, sizeof(sc));
> 	memset(&info, 0, sizeof(info));
> 
> so sc.offscreenbuf = NULL is a no-op.
Oh bugger. My working copy didn't have the allocate commit
d179981b2ddace3dfbf79f828c592a970cd2f009.
Ignore this one then. :)


-- 
Mierswa, Daniel

If you still don't like it, that's ok: that's why I'm boss. I simply know better than you do.
               --- Linus Torvalds, comp.os.linux.advocacy, 1996/07/22

_______________________________________________
barebox mailing list
barebox@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/barebox

^ permalink raw reply	[flat|nested] 10+ messages in thread

* Re: [RESEND PATCH] framebuffer work (especially i.MX)
  2013-01-17  6:32 [RESEND PATCH] framebuffer work (especially i.MX) Daniel Mierswa
                   ` (3 preceding siblings ...)
  2013-01-17  6:32 ` [PATCH 4/4] video/imx: always initialize offscreenbuf member Daniel Mierswa
@ 2013-01-18 10:23 ` Sascha Hauer
  4 siblings, 0 replies; 10+ messages in thread
From: Sascha Hauer @ 2013-01-18 10:23 UTC (permalink / raw)
  To: Daniel Mierswa; +Cc: barebox

On Thu, Jan 17, 2013 at 07:32:55AM +0100, Daniel Mierswa wrote:
> I adjusted the patch series and added the fixes as discussed on the
> mailinglist.
> 
> I still somehow feel that the offscreenbuf member should be set by
> the fd_open routine because I expect fd_close to always work with the
> struct set by fd_open no matter what offscreen parameter the function
> got.
> 
> Daniel Mierswa (4):
>   i.MX21: Add periph. clock register name macros
>   i.MX21/27: don't enable lcd bus clocks too early
>   i.MX27: fix shift amount for PCCR1_PERCLK3_EN

Applied up to this patch.

Thanks
 Sascha

>   video/imx: initialize offscreenbuf member
> 
>  arch/arm/mach-imx/clk-imx21.c               | 67 ++++++++++++++++++++++++++---
>  arch/arm/mach-imx/clk-imx27.c               | 16 ++++---
>  arch/arm/mach-imx/include/mach/imx21-regs.h |  6 ---
>  commands/splash.c                           |  8 ++++
>  drivers/video/imx.c                         | 40 ++++++++++++++---
>  5 files changed, 111 insertions(+), 26 deletions(-)
> 
> -- 
> 1.8.1
> 
> 
> _______________________________________________
> barebox mailing list
> barebox@lists.infradead.org
> http://lists.infradead.org/mailman/listinfo/barebox
> 

-- 
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 |

_______________________________________________
barebox mailing list
barebox@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/barebox

^ permalink raw reply	[flat|nested] 10+ messages in thread

end of thread, other threads:[~2013-01-18 10:23 UTC | newest]

Thread overview: 10+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2013-01-17  6:32 [RESEND PATCH] framebuffer work (especially i.MX) Daniel Mierswa
2013-01-17  6:32 ` [PATCH 1/4] i.MX21: Add periph. clock register name macros Daniel Mierswa
2013-01-17  6:32 ` [PATCH 2/4] i.MX21/27: don't enable lcd bus clocks too early Daniel Mierswa
2013-01-17  6:32 ` [PATCH 3/4] i.MX27: fix shift amount for PCCR1_PERCLK3_EN Daniel Mierswa
2013-01-17  6:32 ` [PATCH 4/4] video/imx: always initialize offscreenbuf member Daniel Mierswa
2013-01-17 11:56   ` Jean-Christophe PLAGNIOL-VILLARD
2013-01-17 13:04     ` Daniel Mierswa
2013-01-17 17:55       ` Sascha Hauer
2013-01-18  4:24         ` Daniel Mierswa
2013-01-18 10:23 ` [RESEND PATCH] framebuffer work (especially i.MX) Sascha Hauer

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox