> From: Brian Norris [mailto:briannorris@xxxxxxxxxxxx] > > > +static inline void > > +rtw_write32s_mask(struct rtw_dev *rtwdev, u32 addr, u32 mask, u32 data) > > +{ > > + BUILD_BUG_ON(addr < 0xC00 || addr >= 0xD00); > > This seems to be a non-traditional use of BUILD_BUG_ON(). Normally, I > see this used for stuff that's guaranteed to be known at compile time -- > structure offsets, constants, etc. This is usually (always?) a constant, > but it passes through a function parameter, so I'm not sure if that's > really guaranteed. > > Anyway...this is failing confusingly for me when I try to build: > > drivers/net/wireless/realtek/rtw88/rtw8822b.c: In function > ‘rtw_write32s_mask’: > drivers/net/wireless/realtek/rtw88/rtw8822b.c:230:176: error: call to > ‘__compiletime_assert_230’ declared with attribute error: BUILD_BUG_ON > failed: addr < 0xC00 || addr >= 0xD00 > BUILD_BUG_ON(addr < 0xC00 || addr >= 0xD00); > > ^ > > I tried to pinpoint which call yielded this, and I think once I deleted > enough calls to rtw_write32s_mask() it came down to this one: > > rtw_write32s_mask(rtwdev, REG_RFEINV, BIT(11) | BIT(10) | 0x3f, > 0x0); > > which doesn't really make sense, as that's a value of 0xcbc. > > What I really think it comes down to is that you can't guarantee > rtw_write32s_mask() will get inlined, and so BUILD_BUG_ON() may not know > what to do with it. Yeah, you're right. I think we should turn it into macro. I will send the fix patch in the next patch set. Thanks! > > I think you should either drop the BUILD_BUG_ON(), or else turn > rtw_write32s_mask() into a macro. > > BTW, I'm using a variant of gcc version 4.9.2. I normally use clang for > kernel builds (as that's the current Chromium OS toolchain), but clang > still regularly has breakage upstream, so I'm still building with an old > GCC toolchain. It's technically still supported in > Documentation/process/changes.rst though! Like you said, should support it anyway :) > > Brian > Yan-Hsuan
