On Mon, 25 May 2020 08:46:37 +0200 Boris Brezillon <boris.brezillon@xxxxxxxxxxxxx> wrote: > On Mon, 25 May 2020 00:13:28 +0200 > Miquel Raynal <miquel.raynal@xxxxxxxxxxx> wrote: > > > Hi Richard, > > > > Richard Weinberger <richard.weinberger@xxxxxxxxx> wrote on Sun, 24 May > > 2020 23:37:13 +0200: > > > > > On Sat, May 9, 2020 at 9:19 PM Miquel Raynal <miquel.raynal@xxxxxxxxxxx> wrote: > > > > ns_debugfs_remove(ns); > > > > - ns_free(ns); /* Free nandsim private resources */ > > > > - nand_release(chip); /* Unregister driver */ > > > > - kfree(ns); /* Free other structures */ > > > > - ns_free_lists(); > > > > + WARN_ON(mtd_device_unregister(nsmtd)); > > > > + ns_free(ns); > > > > + kfree(erase_block_wear); > > > > + nand_cleanup(chip); > > > > + list_for_each_safe(pos, n, &grave_pages) { > > > > + kfree(list_entry(pos, struct grave_page, list)); > > > > + list_del(pos); > > > > > > Are you sure you can use pos after freeing the entry? > > > Smells like use after free. > > > > > > > Mmmmh, I should probably invert those two lines, first call list_del() > > and then call kfree() on list_entry(). > > You can also use list_for_each_entry_safe(): > > struct grave_page *pos, *n; > > ... > > list_for_each_safe(pos, n, &grave_pages, list) { list_for_each_entry_safe(pos, n, &grave_pages, list) { > list_del(&pos->list); > kfree(pos); > } > > > > > Thanks for noticing! > > Miquèl > ______________________________________________________ Linux MTD discussion mailing list http://lists.infradead.org/mailman/listinfo/linux-mtd/