From mboxrd@z Thu Jan 1 00:00:00 1970 From: Dan Carpenter Date: Fri, 28 Feb 2014 13:42:39 +0000 Subject: re: ALSA: hda - Update CA0132 codec to load DSP firmware binary Message-Id: <20140228134238.GA13812@elgon.mountain> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: kernel-janitors@vger.kernel.org [ Not likely a bug. Just confusing code ]. Hello Ian Minett, The patch 01ef7dbffb41: "ALSA: hda - Update CA0132 codec to load DSP firmware binary" from Sep 20, 2012, leads to the following static checker warning: sound/pci/hda/patch_ca0132.c:2281 dspxfr_one_seg() warn: we tested 'hci_write' before and it was 'false' sound/pci/hda/patch_ca0132.c 2255 const struct dsp_image_seg *hci_write = NULL; 2256 unsigned long timeout; 2257 bool dma_active; 2258 2259 if (fls = NULL) 2260 return -EINVAL; 2261 if (is_hci_prog_list_seg(fls)) { 2262 hci_write = fls; 2263 fls = get_next_seg_ptr(fls); 2264 } 2265 2266 if (hci_write && (!fls || is_last(fls))) { If fls->count = 0 and hci_write is non-zero then we return here. 2267 snd_printdd("hci_write\n"); 2268 return dspxfr_hci_write(codec, hci_write); 2269 } 2270 2271 if (fls = NULL || dma_engine = NULL || port_map_mask = 0) { 2272 snd_printdd("Invalid Params\n"); 2273 return -EINVAL; 2274 } 2275 2276 data = fls->data; 2277 chip_addx = fls->chip_addr, 2278 words_to_write = fls->count; words_to_write is fls->count. 2279 2280 if (!words_to_write) 2281 return hci_write ? dspxfr_hci_write(codec, hci_write) : 0; If fls->count is zero then hci_write is zero so the test here is redundant/confusing. Were we intentending to test something else? This code is confusing. 2282 if (reloc) 2283 chip_addx = (chip_addx & (0xFFFF0000 << 2)) + (reloc << 2); regards, dan carpenter