On Thu, Jan 7, 2016 at 11:44 PM, Stephen Hemminger <stephen at networkplumber.org> wrote: > On Thu, 7 Jan 2016 22:03:00 +0530 > Santosh Shukla <sshukla at mvista.com> wrote: > >> #else >> +#if defined(RTE_ARCH_ARM) || defined(RTE_ARCH_ARM64) >> + return 0; /* iopl syscall not supported for ARM/ARM64 */ >> +#endif >> return -1; >> #endif > > Minor net why not: > > #elif defined(RTE_ARCH_ARM) || defined(RTE_ARCH_ARM64) > return -1 > #else > > That way you won't generate two return statements and potentially > trigger warnings from static checkers.
returning -1 would fail for arm/arm64. I guess you meant return 0, right? if so then would need one more return for non-x86/non-arm case. Also I am working on another patchset suggested by Jerin [1] on iopl() in v2 series, That new patchset intended to get rid-off ifdef X_86 clutter for sys/io.h and more iop() definition to arch/platform file. I don't want to include those changes in v3 series as because it seems like two different topic. [1] http://dpdk.org/dev/patchwork/patch/9533/