| Message ID | 20200605161458.2513177-1-hslester96@gmail.com (mailing list archive) |
|---|---|
| State | Superseded, archived |
| Headers | show |
| Series | [v2] fbdev: geode: Add the missed pci_disable_device() in gx1fb_map_video_memory() | expand |
Hi, On 6/5/20 6:14 PM, Chuhong Yuan wrote: > Although gx1fb_probe() has handled the failure of gx1fb_map_video_memory() > partly, it does not call pci_disable_device() as gx1fb_map_video_memory() > calls pci_enable_device(). > Add the missed function call to fix the bug. > > Fixes: 53eed4ec8bcd ("[PATCH] fbdev: geode updates]") This doesn't seem to be a matching commit. The proper commit seems to be: commit a06630f3e7fb29f2524e1d7b009eb8b5a278ba23 Author: Antonino A. Daplas <adaplas@gmail.com> Date: Mon Jun 26 00:27:04 2006 -0700 [PATCH] Detaching fbcon: remove calls to pci_disable_device() Detaching fbcon allows individual drivers to be unloaded. However several drivers call pci_disable_device() upon exit. This function will disable the BAR's which will kill VGA text mode and/or affect X/DRM. To prevent this, remove calls to pci_disable_device() from several drivers. ... which removed pci_disable_device() calls from: drivers/video/aty/radeon_base.c drivers/video/cirrusfb.c drivers/video/geode/gx1fb_core.c drivers/video/geode/gxfb_core.c drivers/video/i810/i810_main.c drivers/video/nvidia/nvidia.c drivers/video/riva/fbdev.c In order to bring back pci_disable_device() calls to gx1fb driver (and other affected ones) you should verify that the issue described in the above commit is no longer present (preferably with testing modified driver on a real hardware). Best regards, -- Bartlomiej Zolnierkiewicz Samsung R&D Institute Poland Samsung Electronics > Signed-off-by: Chuhong Yuan <hslester96@gmail.com> > --- > Changes in v2: > - Fix the typo in the subject. > - Modify the label of error handler. > - Refactor the code. > > drivers/video/fbdev/geode/gx1fb_core.c | 24 +++++++++++++++++------- > 1 file changed, 17 insertions(+), 7 deletions(-) > > diff --git a/drivers/video/fbdev/geode/gx1fb_core.c b/drivers/video/fbdev/geode/gx1fb_core.c > index 5d34d89fb665..15645244e4d0 100644 > --- a/drivers/video/fbdev/geode/gx1fb_core.c > +++ b/drivers/video/fbdev/geode/gx1fb_core.c > @@ -208,29 +208,39 @@ static int gx1fb_map_video_memory(struct fb_info *info, struct pci_dev *dev) > > ret = pci_request_region(dev, 0, "gx1fb (video)"); > if (ret < 0) > - return ret; > + goto err_disable_device; > par->vid_regs = pci_ioremap_bar(dev, 0); > if (!par->vid_regs) > - return -ENOMEM; > + goto err_nomem; > > - if (!request_mem_region(gx_base + 0x8300, 0x100, "gx1fb (display controller)")) > - return -EBUSY; > + if (!request_mem_region(gx_base + 0x8300, 0x100, > + "gx1fb (display controller)")) { > + ret = -EBUSY; > + goto err_disable_device; > + } > par->dc_regs = ioremap(gx_base + 0x8300, 0x100); > if (!par->dc_regs) > - return -ENOMEM; > + goto err_nomem; > > if ((fb_len = gx1_frame_buffer_size()) < 0) > - return -ENOMEM; > + goto err_nomem; > + > info->fix.smem_start = gx_base + 0x800000; > info->fix.smem_len = fb_len; > info->screen_base = ioremap(info->fix.smem_start, info->fix.smem_len); > if (!info->screen_base) > - return -ENOMEM; > + goto err_nomem; > > dev_info(&dev->dev, "%d Kibyte of video memory at 0x%lx\n", > info->fix.smem_len / 1024, info->fix.smem_start); > > return 0; > + > +err_nomem: > + ret = -ENOMEM; > +err_disable_device: > + pci_disable_device(dev); > + return ret; > } > > static int parse_panel_option(struct fb_info *info) >
diff --git a/drivers/video/fbdev/geode/gx1fb_core.c b/drivers/video/fbdev/geode/gx1fb_core.c index 5d34d89fb665..15645244e4d0 100644 --- a/drivers/video/fbdev/geode/gx1fb_core.c +++ b/drivers/video/fbdev/geode/gx1fb_core.c @@ -208,29 +208,39 @@ static int gx1fb_map_video_memory(struct fb_info *info, struct pci_dev *dev) ret = pci_request_region(dev, 0, "gx1fb (video)"); if (ret < 0) - return ret; + goto err_disable_device; par->vid_regs = pci_ioremap_bar(dev, 0); if (!par->vid_regs) - return -ENOMEM; + goto err_nomem; - if (!request_mem_region(gx_base + 0x8300, 0x100, "gx1fb (display controller)")) - return -EBUSY; + if (!request_mem_region(gx_base + 0x8300, 0x100, + "gx1fb (display controller)")) { + ret = -EBUSY; + goto err_disable_device; + } par->dc_regs = ioremap(gx_base + 0x8300, 0x100); if (!par->dc_regs) - return -ENOMEM; + goto err_nomem; if ((fb_len = gx1_frame_buffer_size()) < 0) - return -ENOMEM; + goto err_nomem; + info->fix.smem_start = gx_base + 0x800000; info->fix.smem_len = fb_len; info->screen_base = ioremap(info->fix.smem_start, info->fix.smem_len); if (!info->screen_base) - return -ENOMEM; + goto err_nomem; dev_info(&dev->dev, "%d Kibyte of video memory at 0x%lx\n", info->fix.smem_len / 1024, info->fix.smem_start); return 0; + +err_nomem: + ret = -ENOMEM; +err_disable_device: + pci_disable_device(dev); + return ret; } static int parse_panel_option(struct fb_info *info)
Although gx1fb_probe() has handled the failure of gx1fb_map_video_memory() partly, it does not call pci_disable_device() as gx1fb_map_video_memory() calls pci_enable_device(). Add the missed function call to fix the bug. Fixes: 53eed4ec8bcd ("[PATCH] fbdev: geode updates]") Signed-off-by: Chuhong Yuan <hslester96@gmail.com> --- Changes in v2: - Fix the typo in the subject. - Modify the label of error handler. - Refactor the code. drivers/video/fbdev/geode/gx1fb_core.c | 24 +++++++++++++++++------- 1 file changed, 17 insertions(+), 7 deletions(-)