aboutsummaryrefslogtreecommitdiff
path: root/MAINTAINERS
diff options
context:
space:
mode:
authorLinus Torvalds <torvalds@linux-foundation.org>2019-10-02 16:16:07 -0700
committerLinus Torvalds <torvalds@linux-foundation.org>2019-10-02 16:16:07 -0700
commit0f1a7b3fac0583083ca19d4de47403511ced3521 (patch)
treeda9b1752d94f9ec6f007871972b5679237a6cda7 /MAINTAINERS
parent5021b9182ee805603e3b180220a929af7bd4b960 (diff)
download96b-common-0f1a7b3fac0583083ca19d4de47403511ced3521.tar.gz
timer-of: don't use conditional expression with mixed 'void' types
Randy Dunlap reports on the sparse list that sparse warns about this expression: of_irq->percpu ? free_percpu_irq(of_irq->irq, clkevt) : free_irq(of_irq->irq, clkevt); and honestly, sparse is correct to warn. The return type of free_percpu_irq() is 'void', while free_irq() returns a 'const void *' that is the devname argument passed in to the request_irq(). You can't mix a void type with a non-void types in a conditional expression according to the C standard. It so happens that gcc seems to accept it - and the resulting type of the expression is void - but there's really no reason for the kernel to have this kind of non-standard expression with no real upside. The natural way to write that expression is with an if-statement: if (of_irq->percpu) free_percpu_irq(of_irq->irq, clkevt); else free_irq(of_irq->irq, clkevt); which is more legible anyway. I'm not sure why that timer-of code seems to have this odd pattern. It does the same at allocation time, but at least there the types match, and it makes sense as an expression. Reported-by: Randy Dunlap <rdunlap@infradead.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Diffstat (limited to 'MAINTAINERS')
0 files changed, 0 insertions, 0 deletions