Messages in this thread Patch in this message |  | | Date | Sun, 2 Jun 2019 09:53:47 +1000 (AEST) | From | Finn Thain <> | Subject | Re: [PATCH AUTOSEL 4.4 44/56] video: imsttfb: fix potential NULL pointer dereferences |
| |
On Sat, 1 Jun 2019, Greg Kroah-Hartman wrote:
> On Sat, Jun 01, 2019 at 09:25:48AM -0400, Sasha Levin wrote:
> > From: Kangjie Lu <kjlu@umn.edu> > > > > [ Upstream commit 1d84353d205a953e2381044953b7fa31c8c9702d ] > > ... > > Why only 4.4.y? Shouldn't this be queued up for everything or none? > > thanks, > > greg k-h >
Also, why not check the result of the other ioremap calls? (I should have checked that when this first crossed my inbox...)
From 1d84353d205a953e2381044953b7fa31c8c9702d Mon Sep 17 00:00:00 2001 From: Kangjie Lu <kjlu@umn.edu> Date: Mon, 1 Apr 2019 17:46:58 +0200 Subject: [PATCH] video: imsttfb: fix potential NULL pointer dereferences
In case ioremap fails, the fix releases resources and returns -ENOMEM to avoid NULL pointer dereferences.
Signed-off-by: Kangjie Lu <kjlu@umn.edu> Cc: Aditya Pakki <pakki001@umn.edu> Cc: Finn Thain <fthain@telegraphics.com.au> Cc: Rob Herring <robh@kernel.org> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> [b.zolnierkie: minor patch summary fixup] Signed-off-by: Bartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
diff --git a/drivers/video/fbdev/imsttfb.c b/drivers/video/fbdev/imsttfb.c index 4b9615e4ce74..35bba3c2036d 100644 --- a/drivers/video/fbdev/imsttfb.c +++ b/drivers/video/fbdev/imsttfb.c @@ -1515,6 +1515,11 @@ static int imsttfb_probe(struct pci_dev *pdev, const struct pci_device_id *ent) info->fix.smem_start = addr; info->screen_base = (__u8 *)ioremap(addr, par->ramdac == IBM ? 0x400000 : 0x800000); + if (!info->screen_base) { + release_mem_region(addr, size); + framebuffer_release(info); + return -ENOMEM; + } info->fix.mmio_start = addr + 0x800000; par->dc_regs = ioremap(addr + 0x800000, 0x1000); par->cmap_regs_phys = addr + 0x840000;
|  |