Quoting Paul Walmsley (2019-04-29 22:57:15) > On Mon, 29 Apr 2019, Stephen Boyd wrote: > > > Quoting Paul Walmsley (2019-04-29 12:42:07) > > > On Fri, 26 Apr 2019, Paul Walmsley wrote: > > > > On Fri, 26 Apr 2019, Stephen Boyd wrote: > > > > > > > > > Quoting Paul Walmsley (2019-04-11 01:27:32) > > > > > > Add common library code for the Analog Bits Wide-Range PLL (WRPLL) IP > > > > > > block, as implemented in TSMC CLN28HPC. > > > > > > > > > > I haven't deeply reviewed at all, but I already get two problems when > > > > > compile testing these patches. I can fix them up if nothing else needs > > > > > fixing. > > > > > > > > > > drivers/clk/analogbits/wrpll-cln28hpc.c:165 __wrpll_calc_divq() warn: should 'target_rate << divq' be a 64 bit type? > > > > > drivers/clk/sifive/fu540-prci.c:214:16: error: return expression in void function > > > > > > > > Hmm, that's odd. I will definitely take a look and repost. > > > > > > I'm not able to reproduce these problems. The configs tried here were: > > > > > > - 64-bit RISC-V defconfig w/ PRCI driver enabled (gcc 8.2.0 built with > > > crosstool-NG 1.24.0) > > > > > > - 32-bit ARM defconfig w/ PRCI driver enabled (gcc 8.3.0 built with > > > crosstool-NG 1.24.0) > > > > > > - 32-bit i386 defconfig w/ PRCI driver enabled (gcc > > > 5.4.0-6ubuntu1~16.04.11) > > > > > > Could you post the toolchain and kernel config you're using? > > > > > > > I'm running sparse and smatch too. > > OK. I was able to reproduce the __wrpll_calc_divq() warning. It's been > resolved in the upcoming revision. > > But I don't see the second error with either sparse or smatch. (This is > with sparse at commit 2b96cd804dc7 and smatch at commit f0092daff69d.) > Weird! The return in void function is pretty obvious though so please fix it regardless.
