Message ID | 1360308959-3096-2-git-send-email-agust@denx.de (mailing list archive) |
---|---|
State | New, archived |
Headers | show |
On Fri, 8 Feb 2013 08:35:59 +0100 Anatolij Gustschin <agust@denx.de> wrote: > Since commit f74de500 "drivers/video: fsl-diu-fb: streamline > enabling of interrupts" the interrupt handling in the driver > is broken. Enabling diu interrupt causes an interrupt storm and > results in system lockup. > > The cookie for the interrupt handler function passed to request_irq() > is wrong (it must be a pointer to the diu struct, and not the address > of the pointer to the diu struct). As a result the interrupt handler > can not read diu registers and acknowledge the interrupt. Fix cookie > arguments for request_irq() and free_irq(). > > Registering the diu interrupt handler in probe() must happen before > install_fb() calls since this function registers framebuffer devices > and if fbcon tries to take over framebuffer after registering a frame > buffer device, it will call fb_open of the diu driver and enable the > interrupts. At this time the diu interrupt handler must be registered > already. > > Disabling the interrupts in fsl_diu_release() must happen only if all > other AOIs are closed. Otherwise closing an overlay plane will disable > the interrupts even if the primary frame buffer plane is opened. Add > an appropriate check in the release function. > > ... > > This patch fixes a regression, it should be included in v3.8 since > without it all mpc512x based boards (with DIU support enabled) do not > boot Thanks, I queued both these with a plan to merge into 3.9-rc1. I tagged the patches with "Cc: <stable@vger.kernel.org>" so they should get backported into 3.8.1 and possibly earlier kernels. Sound OK? -- To unsubscribe from this list: send the line "unsubscribe linux-fbdev" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Tue, 12 Feb 2013 16:54:58 -0800
Andrew Morton <akpm@linux-foundation.org> wrote:
> Thanks, I queued both these with a plan to merge into 3.9-rc1.
No I didn't. The patches have already found their way into linux-next
via some other tree.
Without any -stable tags, either. You don't think we should fix the
"24 and 16 bpp" thing in 3.8.x and earlier?
--
To unsubscribe from this list: send the line "unsubscribe linux-fbdev" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
On Tue, 12 Feb 2013 17:01:55 -0800 Andrew Morton <akpm@linux-foundation.org> wrote: > On Tue, 12 Feb 2013 16:54:58 -0800 > Andrew Morton <akpm@linux-foundation.org> wrote: > > > Thanks, I queued both these with a plan to merge into 3.9-rc1. > > No I didn't. The patches have already found their way into linux-next > via some other tree. Yes, via MPC5XXX tree. Since I didn't receive response from fbdev maintainer I applied both patches in my powerpc MPC5XXX next branch so that these can be exposed in linux-next at least. MPC5121 uses the fsl-diu-fb driver, so these patches could go via MPC5XXX tree, but I can't push them via my tree without an ACK from fbdev maintainer. > Without any -stable tags, either. You don't think we should fix the > "24 and 16 bpp" thing in 3.8.x and earlier? I didn't add any -stable tags since my hope was that the patches will go into v3.8 via fbdev tree. It would be good to fix the bpp issue in 3.8.x. But the issue is not critical for earlier maintained stable versions I think. Thanks, Anatolij -- To unsubscribe from this list: send the line "unsubscribe linux-fbdev" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Wed, 13 Feb 2013 10:21:52 +0100 Anatolij Gustschin <agust@denx.de> wrote: > On Tue, 12 Feb 2013 17:01:55 -0800 > Andrew Morton <akpm@linux-foundation.org> wrote: > > > On Tue, 12 Feb 2013 16:54:58 -0800 > > Andrew Morton <akpm@linux-foundation.org> wrote: > > > > > Thanks, I queued both these with a plan to merge into 3.9-rc1. > > > > No I didn't. The patches have already found their way into linux-next > > via some other tree. > > Yes, via MPC5XXX tree. Since I didn't receive response from fbdev > maintainer I applied both patches in my powerpc MPC5XXX next branch > so that these can be exposed in linux-next at least. MPC5121 uses the > fsl-diu-fb driver, so these patches could go via MPC5XXX tree, but I > can't push them via my tree without an ACK from fbdev maintainer. The patches look OK to me - I suggest you go ahead and merge them up for 3.9-rc1. > > Without any -stable tags, either. You don't think we should fix the > > "24 and 16 bpp" thing in 3.8.x and earlier? > > I didn't add any -stable tags since my hope was that the patches > will go into v3.8 via fbdev tree. It would be good to fix the bpp > issue in 3.8.x. But the issue is not critical for earlier maintained > stable versions I think. Please remember to add the cc:stable to the changelogs if you want these in 3.8.x and earlier. -- To unsubscribe from this list: send the line "unsubscribe linux-fbdev" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Wed, 13 Feb 2013 11:55:24 -0800 Andrew Morton <akpm@linux-foundation.org> wrote: > On Wed, 13 Feb 2013 10:21:52 +0100 > Anatolij Gustschin <agust@denx.de> wrote: > > > On Tue, 12 Feb 2013 17:01:55 -0800 > > Andrew Morton <akpm@linux-foundation.org> wrote: > > > > > On Tue, 12 Feb 2013 16:54:58 -0800 > > > Andrew Morton <akpm@linux-foundation.org> wrote: > > > > > > > Thanks, I queued both these with a plan to merge into 3.9-rc1. > > > > > > No I didn't. The patches have already found their way into linux-next > > > via some other tree. > > > > Yes, via MPC5XXX tree. Since I didn't receive response from fbdev > > maintainer I applied both patches in my powerpc MPC5XXX next branch > > so that these can be exposed in linux-next at least. MPC5121 uses the > > fsl-diu-fb driver, so these patches could go via MPC5XXX tree, but I > > can't push them via my tree without an ACK from fbdev maintainer. > > The patches look OK to me - I suggest you go ahead and merge them up > for 3.9-rc1. Okay, will do. > > > Without any -stable tags, either. You don't think we should fix the > > > "24 and 16 bpp" thing in 3.8.x and earlier? > > > > I didn't add any -stable tags since my hope was that the patches > > will go into v3.8 via fbdev tree. It would be good to fix the bpp > > issue in 3.8.x. But the issue is not critical for earlier maintained > > stable versions I think. > > Please remember to add the cc:stable to the changelogs if you want > these in 3.8.x and earlier. Okay, thanks! Anatolij -- To unsubscribe from this list: send the line "unsubscribe linux-fbdev" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
diff --git a/drivers/video/fsl-diu-fb.c b/drivers/video/fsl-diu-fb.c index 1ca32c5..d33e872 100644 --- a/drivers/video/fsl-diu-fb.c +++ b/drivers/video/fsl-diu-fb.c @@ -1232,6 +1232,16 @@ static int fsl_diu_ioctl(struct fb_info *info, unsigned int cmd, return 0; } +static inline void fsl_diu_enable_interrupts(struct fsl_diu_data *data) +{ + u32 int_mask = INT_UNDRUN; /* enable underrun detection */ + + if (IS_ENABLED(CONFIG_NOT_COHERENT_CACHE)) + int_mask |= INT_VSYNC; /* enable vertical sync */ + + clrbits32(&data->diu_reg->int_mask, int_mask); +} + /* turn on fb if count == 1 */ static int fsl_diu_open(struct fb_info *info, int user) @@ -1251,19 +1261,7 @@ static int fsl_diu_open(struct fb_info *info, int user) if (res < 0) mfbi->count--; else { - struct fsl_diu_data *data = mfbi->parent; - -#ifdef CONFIG_NOT_COHERENT_CACHE - /* - * Enable underrun detection and vertical sync - * interrupts. - */ - clrbits32(&data->diu_reg->int_mask, - INT_UNDRUN | INT_VSYNC); -#else - /* Enable underrun detection */ - clrbits32(&data->diu_reg->int_mask, INT_UNDRUN); -#endif + fsl_diu_enable_interrupts(mfbi->parent); fsl_diu_enable_panel(info); } } @@ -1283,9 +1281,18 @@ static int fsl_diu_release(struct fb_info *info, int user) mfbi->count--; if (mfbi->count == 0) { struct fsl_diu_data *data = mfbi->parent; + bool disable = true; + int i; - /* Disable interrupts */ - out_be32(&data->diu_reg->int_mask, 0xffffffff); + /* Disable interrupts only if all AOIs are closed */ + for (i = 0; i < NUM_AOIS; i++) { + struct mfb_info *mi = data->fsl_diu_info[i].par; + + if (mi->count) + disable = false; + } + if (disable) + out_be32(&data->diu_reg->int_mask, 0xffffffff); fsl_diu_disable_panel(info); } @@ -1614,14 +1621,6 @@ static int __devinit fsl_diu_probe(struct platform_device *pdev) out_be32(&data->diu_reg->desc[1], data->dummy_ad.paddr); out_be32(&data->diu_reg->desc[2], data->dummy_ad.paddr); - for (i = 0; i < NUM_AOIS; i++) { - ret = install_fb(&data->fsl_diu_info[i]); - if (ret) { - dev_err(&pdev->dev, "could not register fb %d\n", i); - goto error; - } - } - /* * Older versions of U-Boot leave interrupts enabled, so disable * all of them and clear the status register. @@ -1630,12 +1629,21 @@ static int __devinit fsl_diu_probe(struct platform_device *pdev) in_be32(&data->diu_reg->int_status); ret = request_irq(data->irq, fsl_diu_isr, 0, "fsl-diu-fb", - &data->diu_reg); + data->diu_reg); if (ret) { dev_err(&pdev->dev, "could not claim irq\n"); goto error; } + for (i = 0; i < NUM_AOIS; i++) { + ret = install_fb(&data->fsl_diu_info[i]); + if (ret) { + dev_err(&pdev->dev, "could not register fb %d\n", i); + free_irq(data->irq, data->diu_reg); + goto error; + } + } + sysfs_attr_init(&data->dev_attr.attr); data->dev_attr.attr.name = "monitor"; data->dev_attr.attr.mode = S_IRUGO|S_IWUSR; @@ -1667,7 +1675,7 @@ static int fsl_diu_remove(struct platform_device *pdev) data = dev_get_drvdata(&pdev->dev); disable_lcdc(&data->fsl_diu_info[0]); - free_irq(data->irq, &data->diu_reg); + free_irq(data->irq, data->diu_reg); for (i = 0; i < NUM_AOIS; i++) uninstall_fb(&data->fsl_diu_info[i]);
Since commit f74de500 "drivers/video: fsl-diu-fb: streamline enabling of interrupts" the interrupt handling in the driver is broken. Enabling diu interrupt causes an interrupt storm and results in system lockup. The cookie for the interrupt handler function passed to request_irq() is wrong (it must be a pointer to the diu struct, and not the address of the pointer to the diu struct). As a result the interrupt handler can not read diu registers and acknowledge the interrupt. Fix cookie arguments for request_irq() and free_irq(). Registering the diu interrupt handler in probe() must happen before install_fb() calls since this function registers framebuffer devices and if fbcon tries to take over framebuffer after registering a frame buffer device, it will call fb_open of the diu driver and enable the interrupts. At this time the diu interrupt handler must be registered already. Disabling the interrupts in fsl_diu_release() must happen only if all other AOIs are closed. Otherwise closing an overlay plane will disable the interrupts even if the primary frame buffer plane is opened. Add an appropriate check in the release function. Signed-off-by: Anatolij Gustschin <agust@denx.de> Cc: Timur Tabi <timur@tabi.org> --- This patch fixes a regression, it should be included in v3.8 since without it all mpc512x based boards (with DIU support enabled) do not boot Changes in v2: - don't add can_handle_irq flag and do the interrupt registration before registering frame buffers instead, as pointed by Timur drivers/video/fsl-diu-fb.c | 58 +++++++++++++++++++++++++------------------- 1 files changed, 33 insertions(+), 25 deletions(-)