Ping-Ke Shih <pkshih@xxxxxxxxxxx> writes: > On Sat, 2024-05-04 at 16:38 +0300, Kalle Valo wrote: >> >> Ping-Ke Shih <pkshih@xxxxxxxxxxx> writes: >> >> > On Sat, 2024-05-04 at 07:10 +0000, Kalle Valo wrote: >> > > drivers/net/wireless/realtek/rtlwifi/rtl8192d/hw_common.c:986:2: >> > > ./include/linux/bitfield.h:189:15: warning: 'rate_mask' is used >> > > uninitialized [-Wuninitialized] >> > > >> > >> > I have updated gcc to 13.2.1 (local build), but still can't see this warning. >> >> I'm not sure if you aware that crosstool provides an easy way to test >> different compilers. More info here: >> >> https://wireless.wiki.kernel.org/en/users/drivers/ath10k/codingstyle#tools > > > I didn't know that before, and I use that convenient cross build now. :-) There are also similar packages for clang: https://mirrors.edge.kernel.org/pub/tools/llvm/ > A question about GNUmakefile, it seems miss > > include Makefile > > Otherwise, 'make' told me no rules to make. Indeed, thanks for reporting that. I fixed now the wiki. >> > Then, I force to add -Wuninitialized and delete -Wno-maybe-uninitialized >> > locally, and I can see this warning (but -Wmaybe-uninitialized instead). >> > >> > I did suspect wireless-next tree does more compiler options, so I also >> > tried to merge the tree, but still the same. >> >> Weird, maybe it's due to different kconfig options? I have attached my >> .config which I use to compile test wireless trees. > > Using your .config can reproduce the symptom, and then I try to bisect CONFIG_xxx. > The cause is "CONFIG_DEBUG_SECTION_MISMATCH=y". My .config didn't have that. > > I also tried my original compiler that can also see the warning with > new .config. Ah, I would not have never guessed that option could cause something like this. Very good to know. -- https://patchwork.kernel.org/project/linux-wireless/list/ https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatches
