aboutsummaryrefslogtreecommitdiffstats
path: root/net/core/filter.c
diff options
context:
space:
mode:
authorDaniel Borkmann <daniel@iogearbox.net>2017-01-12 11:51:33 +0100
committerDavid S. Miller <davem@davemloft.net>2017-01-12 10:00:31 -0500
commit62c7989b24dbd348c2507ee6458ebf5637d6ddb5 (patch)
tree352050f1913fae4683e564607374618338c1c773 /net/core/filter.c
parentbpf: pass original insn directly to convert_ctx_access (diff)
downloadlinux-dev-62c7989b24dbd348c2507ee6458ebf5637d6ddb5.tar.xz
linux-dev-62c7989b24dbd348c2507ee6458ebf5637d6ddb5.zip
bpf: allow b/h/w/dw access for bpf's cb in ctx
When structs are used to store temporary state in cb[] buffer that is used with programs and among tail calls, then the generated code will not always access the buffer in bpf_w chunks. We can ease programming of it and let this act more natural by allowing for aligned b/h/w/dw sized access for cb[] ctx member. Various test cases are attached as well for the selftest suite. Potentially, this can also be reused for other program types to pass data around. Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Acked-by: Alexei Starovoitov <ast@kernel.org> Signed-off-by: David S. Miller <davem@davemloft.net>
Diffstat (limited to 'net/core/filter.c')
-rw-r--r--net/core/filter.c41
1 files changed, 33 insertions, 8 deletions
diff --git a/net/core/filter.c b/net/core/filter.c
index 8cfbdefbfb1c..90383860e224 100644
--- a/net/core/filter.c
+++ b/net/core/filter.c
@@ -2776,11 +2776,33 @@ static bool __is_valid_access(int off, int size)
{
if (off < 0 || off >= sizeof(struct __sk_buff))
return false;
+
/* The verifier guarantees that size > 0. */
if (off % size != 0)
return false;
- if (size != sizeof(__u32))
- return false;
+
+ switch (off) {
+ case offsetof(struct __sk_buff, cb[0]) ...
+ offsetof(struct __sk_buff, cb[4]) + sizeof(__u32) - 1:
+ if (size == sizeof(__u16) &&
+ off > offsetof(struct __sk_buff, cb[4]) + sizeof(__u16))
+ return false;
+ if (size == sizeof(__u32) &&
+ off > offsetof(struct __sk_buff, cb[4]))
+ return false;
+ if (size == sizeof(__u64) &&
+ off > offsetof(struct __sk_buff, cb[2]))
+ return false;
+ if (size != sizeof(__u8) &&
+ size != sizeof(__u16) &&
+ size != sizeof(__u32) &&
+ size != sizeof(__u64))
+ return false;
+ break;
+ default:
+ if (size != sizeof(__u32))
+ return false;
+ }
return true;
}
@@ -2799,7 +2821,7 @@ static bool sk_filter_is_valid_access(int off, int size,
if (type == BPF_WRITE) {
switch (off) {
case offsetof(struct __sk_buff, cb[0]) ...
- offsetof(struct __sk_buff, cb[4]):
+ offsetof(struct __sk_buff, cb[4]) + sizeof(__u32) - 1:
break;
default:
return false;
@@ -2823,7 +2845,7 @@ static bool lwt_is_valid_access(int off, int size,
case offsetof(struct __sk_buff, mark):
case offsetof(struct __sk_buff, priority):
case offsetof(struct __sk_buff, cb[0]) ...
- offsetof(struct __sk_buff, cb[4]):
+ offsetof(struct __sk_buff, cb[4]) + sizeof(__u32) - 1:
break;
default:
return false;
@@ -2915,7 +2937,7 @@ static bool tc_cls_act_is_valid_access(int off, int size,
case offsetof(struct __sk_buff, tc_index):
case offsetof(struct __sk_buff, priority):
case offsetof(struct __sk_buff, cb[0]) ...
- offsetof(struct __sk_buff, cb[4]):
+ offsetof(struct __sk_buff, cb[4]) + sizeof(__u32) - 1:
case offsetof(struct __sk_buff, tc_classid):
break;
default:
@@ -3066,8 +3088,11 @@ static u32 sk_filter_convert_ctx_access(enum bpf_access_type type,
si->dst_reg, si->src_reg, insn);
case offsetof(struct __sk_buff, cb[0]) ...
- offsetof(struct __sk_buff, cb[4]):
+ offsetof(struct __sk_buff, cb[4]) + sizeof(__u32) - 1:
BUILD_BUG_ON(FIELD_SIZEOF(struct qdisc_skb_cb, data) < 20);
+ BUILD_BUG_ON((offsetof(struct sk_buff, cb) +
+ offsetof(struct qdisc_skb_cb, data)) %
+ sizeof(__u64));
prog->cb_access = 1;
off = si->off;
@@ -3075,10 +3100,10 @@ static u32 sk_filter_convert_ctx_access(enum bpf_access_type type,
off += offsetof(struct sk_buff, cb);
off += offsetof(struct qdisc_skb_cb, data);
if (type == BPF_WRITE)
- *insn++ = BPF_STX_MEM(BPF_W, si->dst_reg,
+ *insn++ = BPF_STX_MEM(BPF_SIZE(si->code), si->dst_reg,
si->src_reg, off);
else
- *insn++ = BPF_LDX_MEM(BPF_W, si->dst_reg,
+ *insn++ = BPF_LDX_MEM(BPF_SIZE(si->code), si->dst_reg,
si->src_reg, off);
break;