Hi, On Monday, January 09, 2017 01:59:42 PM Sascha Hauer wrote: > On Mon, Jan 02, 2017 at 11:24:02PM +0100, Martin Kaiser wrote: > > The current code calculates the number of color map entries as > > 1 << info->var.bits_per_pixel. For 32bpp modes, 1 << 32 is 0 when > > written to an int variable. As a consequence, the subsequent copying > > of the default (non-empty) color map into our newly allocated color map > > fails and imxfb's probe function returns an error. > > > > On both imx1 and imx21 platforms, the color map is used only for modes > > with <= 8bpp. By allocating 256 entries for the color map, we're on the > > safe side. > > > > Signed-off-by: Martin Kaiser <[email protected]> > > Acked-by: Sascha Hauer <[email protected]>
Thanks, patch queued for 4.11. Best regards, -- Bartlomiej Zolnierkiewicz Samsung R&D Institute Poland Samsung Electronics > Sascha > > > --- > > Re-sending: corrected a typo in the LKML address, sorry for that. > > > > drivers/video/fbdev/imxfb.c | 6 +++++- > > 1 file changed, 5 insertions(+), 1 deletion(-) > > > > diff --git a/drivers/video/fbdev/imxfb.c b/drivers/video/fbdev/imxfb.c > > index fe0c4ee..1b0faad 100644 > > --- a/drivers/video/fbdev/imxfb.c > > +++ b/drivers/video/fbdev/imxfb.c > > @@ -985,7 +985,11 @@ static int imxfb_probe(struct platform_device *pdev) > > */ > > imxfb_check_var(&info->var, info); > > > > - ret = fb_alloc_cmap(&info->cmap, 1 << info->var.bits_per_pixel, 0); > > + /* > > + * For modes > 8bpp, the color map is bypassed. > > + * Therefore, 256 entries are enough. > > + */ > > + ret = fb_alloc_cmap(&info->cmap, 256, 0); > > if (ret < 0) > > goto failed_cmap;
