On 5/23/23 19:38, Markus Elfring wrote:
From: Markus Elfring <elfring@xxxxxxxxxxxxxxxxxxxxx> Date: Tue, 23 May 2023 14:32:39 +0200 The return value was overlooked from a call of the function “fb_alloc_cmap”. * Thus use a corresponding error check. * Add two jump targets so that a bit of exception handling can be better reused at the end of this function. Reported-by: Helge Deller <deller@xxxxxx> Link: https://lore.kernel.org/dri-devel/069f2f78-01f3-9476-d860-2b695c122649@xxxxxx/ Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Fixes: c75f5a550610 ("fbdev: imsttfb: Fix use after free bug in imsttfb_probe") Signed-off-by: Markus Elfring <elfring@xxxxxxxxxxxxxxxxxxxxx> --- drivers/video/fbdev/imsttfb.c | 18 +++++++++++++----- 1 file changed, 13 insertions(+), 5 deletions(-) diff --git a/drivers/video/fbdev/imsttfb.c b/drivers/video/fbdev/imsttfb.c index 975dd682fae4..d3532def4707 100644 --- a/drivers/video/fbdev/imsttfb.c +++ b/drivers/video/fbdev/imsttfb.c @@ -1351,6 +1351,7 @@ static int init_imstt(struct fb_info *info) { struct imstt_par *par = info->par; __u32 i, tmp, *ip, *end; + int ret; tmp = read_reg_le32(par->dc_regs, PRC); if (par->ramdac == IBM) @@ -1419,8 +1420,7 @@ static int init_imstt(struct fb_info *info) if ((info->var.xres * info->var.yres) * (info->var.bits_per_pixel >> 3) > info->fix.smem_len || !(compute_imstt_regvals(par, info->var.xres, info->var.yres))) { printk("imsttfb: %ux%ux%u not supported\n", info->var.xres, info->var.yres, info->var.bits_per_pixel); - framebuffer_release(info); - return -ENODEV; + goto e_nodev; } sprintf(info->fix.id, "IMS TT (%s)", par->ramdac == IBM ? "IBM" : "TVP"); @@ -1452,17 +1452,25 @@ static int init_imstt(struct fb_info *info) FBINFO_HWACCEL_FILLRECT | FBINFO_HWACCEL_YPAN; - fb_alloc_cmap(&info->cmap, 0, 0); + ret = fb_alloc_cmap(&info->cmap, 0, 0); + if (ret) + goto release_framebuffer; if (register_framebuffer(info) < 0) { - framebuffer_release(info); - return -ENODEV; + fb_dealloc_cmap(&info->cmap); + goto e_nodev; } tmp = (read_reg_le32(par->dc_regs, SSTATUS) & 0x0f00) >> 8; fb_info(info, "%s frame buffer; %uMB vram; chip version %u\n", info->fix.id, info->fix.smem_len >> 20, tmp); return 0; + +e_nodev: + ret = -ENODEV;
I think the return value isn't checked at all, so you could simply return below "-ENODEV" for all cases (instead of "return ret"). Then you don't need the e_nodev label and can simplify the flow. Helge
+release_framebuffer: + framebuffer_release(info); + return ret; } static int imsttfb_probe(struct pci_dev *pdev, const struct pci_device_id *ent) -- 2.40.1
