libbpf: Fix bail out from 'ringbuf_process_ring()' on error

commit 6032ebb upstream.

The current code bails out with negative and positive returns.
If the callback returns a positive return code, 'ring_buffer__consume()'
and 'ring_buffer__poll()' will return a spurious number of records
consumed, but mostly important will continue the processing loop.

This patch makes positive returns from the callback a no-op.

Fixes: bf99c93 ("libbpf: Add BPF ring buffer support")
Signed-off-by: Pedro Tammela <pctammela@mojatatu.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20210325150115.138750-1-pctammela@mojatatu.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

tools/lib/bpf/ringbuf.c

@@ -227,7 +227,7 @@ static int ringbuf_process_ring(struct ring* r)
 			if ((len & BPF_RINGBUF_DISCARD_BIT) == 0) {
 				sample = (void *)len_ptr + BPF_RINGBUF_HDR_SZ;
 				err = r->sample_cb(r->ctx, sample, len);
-				if (err) {
+				if (err < 0) {
 					/* update consumer pos and bail out */
 					smp_store_release(r->consumer_pos,
 							  cons_pos);