On Tuesday 08 February 2011 09:30:45 ext-eero.nurkkala@xxxxxxxxx, you wrote: > From: Eero Nurkkala <ext-eero.nurkkala@xxxxxxxxx> > > With HIGHMEM, g_phonet oopses with the following path: > pn_rx_complete() -> pskb_pull() -> ... kmap_skb_frag() and > oops with BUG_ON(in_irq()). Thus, if the gadget is enabled, > it'd better conflict with the HIGHMEM. > > Signed-off-by: Eero Nurkkala <ext-eero.nurkkala@xxxxxxxxx> Acked-by: Rémi Denis-Courmont <remi.denis-courmont@xxxxxxxxx> In the long run, we should probably kludge the skb code like we did in cdc-phonet.c. Or better yet, fix kmap_skb_frag() to be less pessimistic. Thing is, the BUG_ON is really bogus here. The way we allocated the skb, the fragment really cannot be in HIGHMEM... -- Rémi Denis-Courmont Nokia Devices R&D, Maemo Software, Helsinki -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
