path: root/icmp.h
diff options
authorStefano Brivio <sbrivio@redhat.com>2021-04-23 22:22:37 +0200
committerStefano Brivio <sbrivio@redhat.com>2021-04-23 22:22:37 +0200
commit38b50dba4704856194ac02b98e492d2349d64058 (patch)
treeb7f7a12479aa53f9e7ac2cae7a4e709ce78d99a5 /icmp.h
parent962bc97cf116519bd11b7e8beeda6dcce033d537 (diff)
passt: Spare some syscalls, add some optimisations from profiling
Avoid a bunch of syscalls on forwarding paths by: - storing minimum and maximum file descriptor numbers for each protocol, fall back to SO_PROTOCOL query only on overlaps - allocating a larger receive buffer -- this can result in more coalesced packets than sendmmsg() can take (UIO_MAXIOV, i.e. 1024), so make sure we don't exceed that within a single call to protocol tap handlers - nesting the handling loop in tap_handler() in the receive loop, so that we have better chances of filling our receive buffer in fewer calls - skipping the recvfrom() in the UDP handler on EPOLLERR -- there's nothing to be done in that case and while at it: - restore the 20ms timer interval for periodic (TCP) events, I accidentally changed that to 100ms in an earlier commit - attempt using SO_ZEROCOPY for UDP -- if it's not available, sendmmsg() will succeed anyway - fix the handling of the status code from sendmmsg(), if it fails, we'll try to discard the first message, hence return 1 from the UDP handler Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
Diffstat (limited to 'icmp.h')
1 files changed, 4 insertions, 0 deletions
diff --git a/icmp.h b/icmp.h
index 9d26050..1941028 100644
--- a/icmp.h
+++ b/icmp.h
@@ -12,10 +12,14 @@ int icmp_sock_init(struct ctx *c);
* struct icmp_ctx - Execution context for ICMP routines
* @s4: ICMP socket number
* @s6: ICMPv6 socket number
+ * @fd_min: Lowest file descriptor number for ICMP/ICMPv6 ever used
+ * @fd_max: Highest file descriptor number for ICMP/ICMPv6 ever used
struct icmp_ctx {
int s4;
int s6;
+ int fd_min;
+ int fd_max;
#endif /* ICMP_H */