Fix bpf obj get() behavior.
#3238
Conversation
gtrevi
requested review from
dthaler,
poornagmsft,
Alan-Jowett,
saxena-anurag,
shankarseal,
dv-msft,
shpalani,
matthewige,
mtfriesen and
rectified95
as code owners
| } | ||
|
|
||
| ebpf_assert(reply.header.id == ebpf_operation_id_t::EBPF_OPERATION_GET_PINNED_OBJECT); | ||
|
|
||
| ebpf_handle_t handle = reply.handle; | ||
| fd_t fd = _create_file_descriptor_for_handle(handle); | ||
| if (fd == ebpf_fd_invalid) { | ||
| *fd = _create_file_descriptor_for_handle(handle); |
There was a problem hiding this comment.
I believe _open_osfhandle sets errno. Hence you could update the _create_file_descriptor_for_handle() prototype to return the EBPF error rather than simply assuming it's EBPF_NO_MEMORY.
There was a problem hiding this comment.
For coherence, I preserved the same behavior for all _create_file_descriptor_for_handle calls across the rest of the code, all returning EBPF_NO_MEMORY.
There was a problem hiding this comment.
I looked at https://learn.microsoft.com/en-us/cpp/c-runtime-library/reference/open-osfhandle?view=msvc-170, but was not able to find if _open_osfhandle() sets errno.
If _open_osfhandle() does set errno, maybe we can file an issue to fix it all the places?
There was a problem hiding this comment.
Hmm. libuv/libuv#1420 reported that back in 2017, it didn't.
Fixes #3234
Description
Fixes the behavior of
bpf_obj_get()to comply with the official specs: https://docs.kernel.org/userspace-api/ebpf/syscall.htmlTesting
CI/CD
Documentation
n.a.
Installation
n.a.