* Re: [linux-next:master 11929/12398] sound/soc/codecs/nau8325.c:430:13: warning: variable 'n2_max' is uninitialized when used here
2025-11-27 5:50 [linux-next:master 11929/12398] sound/soc/codecs/nau8325.c:430:13: warning: variable 'n2_max' is uninitialized when used here kernel test robot
@ 2025-12-01 19:10 ` Nathan Chancellor
0 siblings, 0 replies; 2+ messages in thread
From: Nathan Chancellor @ 2025-12-01 19:10 UTC (permalink / raw)
To: Seven Lee, Jaroslav Kysela, Mark Brown
Cc: llvm, oe-kbuild-all, kernel test robot
On Thu, Nov 27, 2025 at 01:50:44PM +0800, kernel test robot wrote:
> Hi Jaroslav,
>
> First bad commit (maybe != root cause):
Indeed, this appears to be a problem from commit c0a3873b9938 ("ASoC:
nau8325: new driver"), it was simply not visible before the driver was
actually hooked into the build system.
> tree: https://git.kernel.org/pub/scm/linux/kernel/git/next/linux-next.git master
> head: ef68bf704646690aba5e81c2f7be8d6ef13d7ad8
> commit: cd41d3420ef658b2ca902d7677536ec8e25b610a [11929/12398] ASoC: nau8325: add missing build config
> config: loongarch-allmodconfig (https://download.01.org/0day-ci/archive/20251127/202511271353.g6GYVUt6-lkp@intel.com/config)
> compiler: clang version 19.1.7 (https://github.com/llvm/llvm-project cd708029e0b2869e80abe31ddb175f7c35361f90)
> reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20251127/202511271353.g6GYVUt6-lkp@intel.com/reproduce)
>
> If you fix the issue in a separate patch/commit (i.e. not just a new version of
> the same patch/commit), kindly add following tags
> | Reported-by: kernel test robot <lkp@intel.com>
> | Closes: https://lore.kernel.org/oe-kbuild-all/202511271353.g6GYVUt6-lkp@intel.com/
>
> All warnings (new ones prefixed by >>):
>
> >> sound/soc/codecs/nau8325.c:430:13: warning: variable 'n2_max' is uninitialized when used here [-Wuninitialized]
> 430 | *n2_sel = n2_max;
> | ^~~~~~
> sound/soc/codecs/nau8325.c:389:52: note: initialize the variable 'n2_max' to silence this warning
> 389 | int i, j, mclk, mclk_max, ratio, ratio_sel, n2_max;
> | ^
> | = 0
> >> sound/soc/codecs/nau8325.c:431:11: warning: variable 'ratio_sel' is uninitialized when used here [-Wuninitialized]
> 431 | ratio = ratio_sel;
> | ^~~~~~~~~
> sound/soc/codecs/nau8325.c:389:44: note: initialize the variable 'ratio_sel' to silence this warning
> 389 | int i, j, mclk, mclk_max, ratio, ratio_sel, n2_max;
> | ^
> | = 0
> 2 warnings generated.
This warning seems legitimate even though it seems like there is no harm
at runtime?
> vim +/n2_max +430 sound/soc/codecs/nau8325.c
>
> c0a3873b9938bf Seven Lee 2024-03-27 384
> c0a3873b9938bf Seven Lee 2024-03-27 385 static int nau8325_clksrc_choose(struct nau8325 *nau8325,
> c0a3873b9938bf Seven Lee 2024-03-27 386 const struct nau8325_srate_attr **srate_table,
> c0a3873b9938bf Seven Lee 2024-03-27 387 int *n1_sel, int *mult_sel, int *n2_sel)
> c0a3873b9938bf Seven Lee 2024-03-27 388 {
> c0a3873b9938bf Seven Lee 2024-03-27 389 int i, j, mclk, mclk_max, ratio, ratio_sel, n2_max;
> c0a3873b9938bf Seven Lee 2024-03-27 390
> c0a3873b9938bf Seven Lee 2024-03-27 391 if (!nau8325->mclk || !nau8325->fs)
> c0a3873b9938bf Seven Lee 2024-03-27 392 goto proc_err;
> c0a3873b9938bf Seven Lee 2024-03-27 393
> c0a3873b9938bf Seven Lee 2024-03-27 394 /* select sampling rate and MCLK_SRC */
> c0a3873b9938bf Seven Lee 2024-03-27 395 *srate_table = target_srate_attribute(nau8325->fs);
> c0a3873b9938bf Seven Lee 2024-03-27 396 if (!*srate_table)
> c0a3873b9938bf Seven Lee 2024-03-27 397 goto proc_err;
> c0a3873b9938bf Seven Lee 2024-03-27 398
> c0a3873b9938bf Seven Lee 2024-03-27 399 /* First check clock from MCLK directly, decide N2 for MCLK_SRC.
> c0a3873b9938bf Seven Lee 2024-03-27 400 * If not good, consider 1/N1 and Multiplier.
> c0a3873b9938bf Seven Lee 2024-03-27 401 */
> c0a3873b9938bf Seven Lee 2024-03-27 402 ratio = nau8325_clksrc_n2(nau8325, *srate_table, nau8325->mclk, n2_sel);
> c0a3873b9938bf Seven Lee 2024-03-27 403 if (ratio != NAU8325_MCLK_FS_RATIO_NUM) {
> c0a3873b9938bf Seven Lee 2024-03-27 404 *n1_sel = 0;
> c0a3873b9938bf Seven Lee 2024-03-27 405 *mult_sel = CLK_PROC_BYPASS;
> c0a3873b9938bf Seven Lee 2024-03-27 406 *n2_sel = MCLK_SRC;
> c0a3873b9938bf Seven Lee 2024-03-27 407 goto proc_done;
> c0a3873b9938bf Seven Lee 2024-03-27 408 }
> c0a3873b9938bf Seven Lee 2024-03-27 409
> c0a3873b9938bf Seven Lee 2024-03-27 410 /* Get MCLK_SRC through 1/N, Multiplier, and then 1/N2. */
> c0a3873b9938bf Seven Lee 2024-03-27 411 mclk_max = 0;
> c0a3873b9938bf Seven Lee 2024-03-27 412 for (i = 0; i < ARRAY_SIZE(mclk_n1_div); i++) {
> c0a3873b9938bf Seven Lee 2024-03-27 413 for (j = 0; j < ARRAY_SIZE(mclk_n3_mult); j++) {
> c0a3873b9938bf Seven Lee 2024-03-27 414 mclk = nau8325->mclk << mclk_n3_mult[j].param;
> c0a3873b9938bf Seven Lee 2024-03-27 415 mclk = mclk / mclk_n1_div[i].param;
> c0a3873b9938bf Seven Lee 2024-03-27 416 ratio = nau8325_clksrc_n2(nau8325,
> c0a3873b9938bf Seven Lee 2024-03-27 417 *srate_table, mclk, n2_sel);
> c0a3873b9938bf Seven Lee 2024-03-27 418 if (ratio != NAU8325_MCLK_FS_RATIO_NUM &&
> c0a3873b9938bf Seven Lee 2024-03-27 419 (mclk_max < mclk || i > *n1_sel)) {
> c0a3873b9938bf Seven Lee 2024-03-27 420 mclk_max = mclk;
> c0a3873b9938bf Seven Lee 2024-03-27 421 n2_max = *n2_sel;
> c0a3873b9938bf Seven Lee 2024-03-27 422 *n1_sel = i;
> c0a3873b9938bf Seven Lee 2024-03-27 423 *mult_sel = j;
> c0a3873b9938bf Seven Lee 2024-03-27 424 ratio_sel = ratio;
> c0a3873b9938bf Seven Lee 2024-03-27 425 goto proc_done;
n2_max and ratio_sel are only assigned in this if block, which ends in
an unconditional jump to proc_done...
> c0a3873b9938bf Seven Lee 2024-03-27 426 }
> c0a3873b9938bf Seven Lee 2024-03-27 427 }
> c0a3873b9938bf Seven Lee 2024-03-27 428 }
> c0a3873b9938bf Seven Lee 2024-03-27 429 if (mclk_max) {
> c0a3873b9938bf Seven Lee 2024-03-27 @430 *n2_sel = n2_max;
> c0a3873b9938bf Seven Lee 2024-03-27 @431 ratio = ratio_sel;
> c0a3873b9938bf Seven Lee 2024-03-27 432 goto proc_done;
> c0a3873b9938bf Seven Lee 2024-03-27 433 }
which will always skip this if block, so if this if statement is taken,
these values are genuinely uninitialized.
However, as far as I can tell, if the first if statement is not taken,
mclk_max will be 0, so it seems like this if statement is never taken
and is dead code? At the same time, won't 'mclk_max < mclk' always be
true unless mclk itself is 0, so what purpose does mclk_max serve
otherwise? Was it intentional that the first if statement ends in a jump
to proc_done? It looks like the value of n2_max is never actually used
outside of this dead if statement.
> c0a3873b9938bf Seven Lee 2024-03-27 435 proc_err:
> c0a3873b9938bf Seven Lee 2024-03-27 436 dev_dbg(nau8325->dev, "The MCLK %d is invalid. It can't get MCLK_SRC of 256/400/500 FS (%d)",
> c0a3873b9938bf Seven Lee 2024-03-27 437 nau8325->mclk, nau8325->fs);
> c0a3873b9938bf Seven Lee 2024-03-27 438 return -EINVAL;
> c0a3873b9938bf Seven Lee 2024-03-27 439 proc_done:
> c0a3873b9938bf Seven Lee 2024-03-27 440 dev_dbg(nau8325->dev, "nau8325->fs=%d,range=0x%x, %s, (n1,mu,n2,dmu):(%d,%d,%d), MCLK_SRC=%uHz (%d)",
> c0a3873b9938bf Seven Lee 2024-03-27 441 nau8325->fs, (*srate_table)->range,
> c0a3873b9938bf Seven Lee 2024-03-27 442 (*srate_table)->max ? "MAX" : "MIN",
> c0a3873b9938bf Seven Lee 2024-03-27 443 *n1_sel == CLK_PROC_BYPASS ?
> c0a3873b9938bf Seven Lee 2024-03-27 444 CLK_PROC_BYPASS : mclk_n1_div[*n1_sel].param,
> c0a3873b9938bf Seven Lee 2024-03-27 445 *mult_sel == CLK_PROC_BYPASS ?
> c0a3873b9938bf Seven Lee 2024-03-27 446 CLK_PROC_BYPASS : 1 << mclk_n3_mult[*mult_sel].param,
> c0a3873b9938bf Seven Lee 2024-03-27 447 1 << mclk_n2_div[*n2_sel].param,
> c0a3873b9938bf Seven Lee 2024-03-27 448 (*srate_table)->mclk_src[ratio],
> c0a3873b9938bf Seven Lee 2024-03-27 449 (*srate_table)->mclk_src[ratio] / nau8325->fs);
> c0a3873b9938bf Seven Lee 2024-03-27 450
> c0a3873b9938bf Seven Lee 2024-03-27 451 return 0;
> c0a3873b9938bf Seven Lee 2024-03-27 452 }
> c0a3873b9938bf Seven Lee 2024-03-27 453
Cheers,
Nathan
^ permalink raw reply [flat|nested] 2+ messages in thread