pipe2() with flags=0 is idential to pipe(), but the pipe() system call
abuses the fact that system calls return two registers to avoid the use of
copyout() making it hard to trace with dtrace (since the kernel ignores
the arguments) and requring a machine dependent assembly wrapper in libc
for each version.
Details
Details
Diff Detail
Diff Detail
- Repository
- rS FreeBSD src repository - subversion
- Lint
Lint Passed - Unit
No Test Coverage - Build Status
Buildable 4229 Build 4273: arc lint + arc unit
Event Timeline
Comment Actions
Does pipe.o need to be added to NOASM, or does t figure this out when pipe.c is in SRCS?
Comment Actions
It looks like NOASM wasn't required in my test tree because D6816 was applied. Now fixed.
Comment Actions
I'd like to squeeze it in for 11 so aarch64 and riscv can not support the syscall, but it's not critical. I need to test the result more thoroughly before I can propose it.