On 2023/8/17 14:13, John Fastabend wrote:
Liu Jian wrote:
If the sockmap msg redirection function is used only to forward packets
and no other operation, the execution result of the BPF_SK_MSG_VERDICT
program is the same each time. In this case, the BPF program only needs to
be run once. Add BPF_F_PERMANENTLY flag to bpf_msg_redirect_map() and
bpf_msg_redirect_hash() to implement this ability.
I like the use case. Did you consider using
long bpf_msg_apply_bytes(struct sk_msg_buff *msg, u32 bytes)
This could be set to UINT32_MAX and then the BPF prog would only be run
every 0xfffffff bytes.
I didn't realize that this feature could be used for this, and I thought
it should have the same effect. Thanks John.
Then we can enable this function in the bpf program as follows:
bpf_msg_redirect_hash(xx, xx, xx, BPF_F_INGRESS | BPF_F_PERMANENTLY);
Test results using netperf TCP_STREAM mode:
for i in 1 64 128 512 1k 2k 32k 64k 100k 500k 1m;then
netperf -T 1,2 -t TCP_STREAM -H 127.0.0.1 -l 20 -- -m $i -s 100m,100m -S 100m,100m
done
before:
3.84 246.52 496.89 1885.03 3415.29 6375.03 40749.09 48764.40 51611.34 55678.26 55992.78
after:
4.43 279.20 555.82 2080.79 3870.70 7105.44 41836.41 49709.75 51861.56 55211.00 54566.85
I suspect comparing against
bpf_msg_redirect_hash(...)
bpf_msg_apply_bytes(msg, UINT32_MAX)
the diff will be rather small. I agree the API is nicer though to simply
Yes, it should have the same effect and looks good to me.
set the flag. Its too bad we didn't think to add a forever to apply_bytes.
I would prefer this API for example,
bpf_msg_redirect_hash(...)
bpf_msg_apply_bytes(msg, 0, PERMANENT);
What do you mean by this? Should I post another version for this?
Given we have apply_bytes is it still useful to have a PERMANENT flag
in your use case? Here we would just reset to UNINT32_MAX if we reached
max bytes.
If apply_bytes is set to UNINT32_MAX, the number of times that the bpf
program runs should be small enough to meet my needs.
Signed-off-by: Liu Jian <liujian56@xxxxxxxxxx>
---
include/linux/skmsg.h | 1 +
include/uapi/linux/bpf.h | 7 +++++--
net/core/skmsg.c | 1 +
net/core/sock_map.c | 4 ++--
net/ipv4/tcp_bpf.c | 21 +++++++++++++++------
tools/include/uapi/linux/bpf.h | 7 +++++--
6 files changed, 29 insertions(+), 12 deletions(-)
[...]
diff --git a/tools/include/uapi/linux/bpf.h b/tools/include/uapi/linux/bpf.h
index 70da85200695..cf622ea4f018 100644
--- a/tools/include/uapi/linux/bpf.h
+++ b/tools/include/uapi/linux/bpf.h
@@ -3004,7 +3004,8 @@ union bpf_attr {
* egress interfaces can be used for redirection. The
* **BPF_F_INGRESS** value in *flags* is used to make the
* distinction (ingress path is selected if the flag is present,
- * egress path otherwise). This is the only flag supported for now.
+ * egress path otherwise). The **BPF_F_PERMANENTLY** value in
+ * *flags* is used to indicates whether the eBPF result is permanent.
We at least need to document what happens if PERMANENTLY and apply_bytes are
used together.
* Return
* **SK_PASS** on success, or **SK_DROP** on error.
*
