Hi Dan,
This is a counter width check. The counter widths can be 8,16,32,64 bits.
That check is a bug and not dead code, I am pushing a patch for that.
Warm regards,
PK
On Thu, Aug 15, 2024 at 2:57 AM Dan Carpenter <dan.carpenter@xxxxxxxxxx> wrote:
>
> Hello Pavitrakumar M,
>
> Commit c8981d9230d8 ("crypto: spacc - Add SPAcc Skcipher support")
> from Jul 29, 2024 (linux-next), leads to the following Smatch static
> checker warning:
>
> drivers/crypto/dwc-spacc/spacc_skcipher.c:458 spacc_cipher_process()
> warn: bitwise AND condition is false here
>
> drivers/crypto/dwc-spacc/spacc_skcipher.c
> 441 } else if (readl(device_h->regmap + SPACC_REG_VERSION_EXT_3)
> 442 & (0x1)) { /* 16-bit counter width */
> 443
> 444 for (i = 14; i < 16; i++) {
> 445 num_iv <<= 8;
> 446 num_iv |= ivc1[i];
> 447 }
> 448
> 449 diff = SPACC_CTR_IV_MAX16 - num_iv;
> 450
> 451 if (len > diff) {
> 452 name = salg->calg->cra_name;
> 453 ret = spacc_skcipher_fallback(name,
> 454 req, enc_dec);
> 455 return ret;
> 456 }
> 457 } else if (readl(device_h->regmap + SPACC_REG_VERSION_EXT_3)
> --> 458 & (0x0)) { /* 8-bit counter width */
> ^^^^^^^
> What the deal here? Generally in the kernel we don't allow dead code. If it's
> necessary then we can add it later.
>
> 459
> 460 for (i = 15; i < 16; i++) {
> 461 num_iv <<= 8;
> 462 num_iv |= ivc1[i];
> 463 }
> 464
> 465 diff = SPACC_CTR_IV_MAX8 - num_iv;
> 466
> 467 if (len > diff) {
> 468 name = salg->calg->cra_name;
> 469 ret = spacc_skcipher_fallback(name,
> 470 req, enc_dec);
> 471 return ret;
> 472 }
> 473 }
> 474 }
>
> regards,
> dan carpenter
