diff mbox series

[bpf-next,v4] bpf, test_run: fix &xdp_frame misplacement for LIVE_FRAMES

Message ID 20230215185440.4126672-1-aleksander.lobakin@intel.com
State New
Headers
Series [bpf-next,v4] bpf, test_run: fix &xdp_frame misplacement for LIVE_FRAMES |

Commit Message

Alexander Lobakin Feb. 15, 2023, 6:54 p.m. UTC 
  &xdp_buff and &xdp_frame are bound in a way that

xdp_buff->data_hard_start == xdp_frame

It's always the case and e.g. xdp_convert_buff_to_frame() relies on
this.
IOW, the following:

	for (u32 i = 0; i < 0xdead; i++) {
		xdpf = xdp_convert_buff_to_frame(&xdp);
		xdp_convert_frame_to_buff(xdpf, &xdp);
	}

shouldn't ever modify @xdpf's contents or the pointer itself.
However, "live packet" code wrongly treats &xdp_frame as part of its
context placed *before* the data_hard_start. With such flow,
data_hard_start is sizeof(*xdpf) off to the right and no longer points
to the XDP frame.

Instead of replacing `sizeof(ctx)` with `offsetof(ctx, xdpf)` in several
places and praying that there are no more miscalcs left somewhere in the
code, unionize ::frm with ::data in a flex array, so that both starts
pointing to the actual data_hard_start and the XDP frame actually starts
being a part of it, i.e. a part of the headroom, not the context.
A nice side effect is that the maximum frame size for this mode gets
increased by 40 bytes, as xdp_buff::frame_sz includes everything from
data_hard_start (-> includes xdpf already) to the end of XDP/skb shared
info.
Also update %MAX_PKT_SIZE accordingly in the selftests code. Leave it
hardcoded for 64 bit && 4k pages, it can be made more flexible later on.

Minor: align `&head->data` with how `head->frm` is assigned for
consistency.
Minor #2: rename 'frm' to 'frame' in &xdp_page_head while at it for
clarity.

(was found while testing XDP traffic generator on ice, which calls
 xdp_convert_frame_to_buff() for each XDP frame)

Fixes: b530e9e1063e ("bpf: Add "live packet" mode for XDP in BPF_PROG_RUN")
Acked-by: Toke Høiland-Jørgensen <toke@redhat.com>
Signed-off-by: Alexander Lobakin <aleksander.lobakin@intel.com>
---
From v3[0]:
- target bpf-next to account adding support for s390 (Martin);
- update both kernel and userspace %MAX_PKT_SIZE accordingly.

From v2[1]:
- update %MAX_PKT_SIZE in the selftests (Daniel, CI bots);
- add conditional static assert to avoid facing the same issue in future;
- pick one Acked-by (Toke).

From v1[2]:
- align `&head->data` with how `head->frm` is assigned for consistency
  (Toke);
- rename 'frm' to 'frame' in &xdp_page_head (Jakub);
- no functional changes.

[0] https://lore.kernel.org/bpf/20230215152141.3753548-1-aleksander.lobakin@intel.com
[1] https://lore.kernel.org/bpf/20230213142747.3225479-1-alexandr.lobakin@intel.com
[2] https://lore.kernel.org/bpf/20230209172827.874728-1-alexandr.lobakin@intel.com
---
 net/bpf/test_run.c                            | 29 +++++++++++++++----
 .../bpf/prog_tests/xdp_do_redirect.c          |  7 +++--
 2 files changed, 27 insertions(+), 9 deletions(-)

Comments

patchwork-bot+netdevbpf@kernel.org Feb. 16, 2023, 1:50 a.m. UTC | #1 
Hello:

This patch was applied to bpf/bpf-next.git (master)
by Martin KaFai Lau <martin.lau@kernel.org>:

On Wed, 15 Feb 2023 19:54:40 +0100 you wrote:
> &xdp_buff and &xdp_frame are bound in a way that
> 
> xdp_buff->data_hard_start == xdp_frame
> 
> It's always the case and e.g. xdp_convert_buff_to_frame() relies on
> this.
> IOW, the following:
> 
> [...]

Here is the summary with links:
  - [bpf-next,v4] bpf, test_run: fix &xdp_frame misplacement for LIVE_FRAMES
    https://git.kernel.org/bpf/bpf-next/c/6c20822fada1

You are awesome, thank you!
Martin KaFai Lau Feb. 17, 2023, 8:32 p.m. UTC | #2 
On 2/15/23 10:54 AM, Alexander Lobakin wrote:
> +#if BITS_PER_LONG == 64 && PAGE_SIZE == SZ_4K
> +/* tools/testing/selftests/bpf/prog_tests/xdp_do_redirect.c:%MAX_PKT_SIZE
> + * must be updated accordingly when any of these changes, otherwise BPF
> + * selftests will fail.
> + */
> +#ifdef __s390x__
> +#define TEST_MAX_PKT_SIZE 3216
> +#else
> +#define TEST_MAX_PKT_SIZE 3408

I have to revert this patch for now. It is not right to assume cache line size:
https://lore.kernel.org/bpf/50c35055-afa9-d01e-9a05-ea5351280e4f@intel.com/

Please resubmit and consider if this static_assert is really needed in the 
kernel test_run.c.

> +#endif
> +static_assert(SKB_WITH_OVERHEAD(TEST_XDP_FRAME_SIZE - XDP_PACKET_HEADROOM) ==
> +	      TEST_MAX_PKT_SIZE);
> +#endif
Alexander Lobakin Feb. 20, 2023, 3:48 p.m. UTC | #3 
From: Martin Kafai Lau <martin.lau@linux.dev>
Date: Fri, 17 Feb 2023 12:32:36 -0800

> On 2/15/23 10:54 AM, Alexander Lobakin wrote:
>> +#if BITS_PER_LONG == 64 && PAGE_SIZE == SZ_4K
>> +/*
>> tools/testing/selftests/bpf/prog_tests/xdp_do_redirect.c:%MAX_PKT_SIZE
>> + * must be updated accordingly when any of these changes, otherwise BPF
>> + * selftests will fail.
>> + */
>> +#ifdef __s390x__
>> +#define TEST_MAX_PKT_SIZE 3216
>> +#else
>> +#define TEST_MAX_PKT_SIZE 3408
> 
> I have to revert this patch for now. It is not right to assume cache
> line size:

Userspace part tries to guess it :D Anyway, let's keep as it was for now.
Sent v5 without the static assertion, I hope it can still hit current
cycle? Or too late already?

> https://lore.kernel.org/bpf/50c35055-afa9-d01e-9a05-ea5351280e4f@intel.com/
> 
> Please resubmit and consider if this static_assert is really needed in
> the kernel test_run.c.
> 
>> +#endif
>> +static_assert(SKB_WITH_OVERHEAD(TEST_XDP_FRAME_SIZE -
>> XDP_PACKET_HEADROOM) ==
>> +          TEST_MAX_PKT_SIZE);
>> +#endif
> 

Thanks,
Olek
diff mbox series

Patch

diff --git a/net/bpf/test_run.c b/net/bpf/test_run.c
index b766a84c8536..1ab396a2b87f 100644
--- a/net/bpf/test_run.c
+++ b/net/bpf/test_run.c
@@ -97,8 +97,11 @@  static bool bpf_test_timer_continue(struct bpf_test_timer *t, int iterations,
 struct xdp_page_head {
 	struct xdp_buff orig_ctx;
 	struct xdp_buff ctx;
-	struct xdp_frame frm;
-	u8 data[];
+	union {
+		/* ::data_hard_start starts here */
+		DECLARE_FLEX_ARRAY(struct xdp_frame, frame);
+		DECLARE_FLEX_ARRAY(u8, data);
+	};
 };
 
 struct xdp_test_data {
@@ -116,6 +119,20 @@  struct xdp_test_data {
 #define TEST_XDP_FRAME_SIZE (PAGE_SIZE - sizeof(struct xdp_page_head))
 #define TEST_XDP_MAX_BATCH 256
 
+#if BITS_PER_LONG == 64 && PAGE_SIZE == SZ_4K
+/* tools/testing/selftests/bpf/prog_tests/xdp_do_redirect.c:%MAX_PKT_SIZE
+ * must be updated accordingly when any of these changes, otherwise BPF
+ * selftests will fail.
+ */
+#ifdef __s390x__
+#define TEST_MAX_PKT_SIZE 3216
+#else
+#define TEST_MAX_PKT_SIZE 3408
+#endif
+static_assert(SKB_WITH_OVERHEAD(TEST_XDP_FRAME_SIZE - XDP_PACKET_HEADROOM) ==
+	      TEST_MAX_PKT_SIZE);
+#endif
+
 static void xdp_test_run_init_page(struct page *page, void *arg)
 {
 	struct xdp_page_head *head = phys_to_virt(page_to_phys(page));
@@ -132,8 +149,8 @@  static void xdp_test_run_init_page(struct page *page, void *arg)
 	headroom -= meta_len;
 
 	new_ctx = &head->ctx;
-	frm = &head->frm;
-	data = &head->data;
+	frm = head->frame;
+	data = head->data;
 	memcpy(data + headroom, orig_ctx->data_meta, frm_len);
 
 	xdp_init_buff(new_ctx, TEST_XDP_FRAME_SIZE, &xdp->rxq);
@@ -223,7 +240,7 @@  static void reset_ctx(struct xdp_page_head *head)
 	head->ctx.data = head->orig_ctx.data;
 	head->ctx.data_meta = head->orig_ctx.data_meta;
 	head->ctx.data_end = head->orig_ctx.data_end;
-	xdp_update_frame_from_buff(&head->ctx, &head->frm);
+	xdp_update_frame_from_buff(&head->ctx, head->frame);
 }
 
 static int xdp_recv_frames(struct xdp_frame **frames, int nframes,
@@ -285,7 +302,7 @@  static int xdp_test_run_batch(struct xdp_test_data *xdp, struct bpf_prog *prog,
 		head = phys_to_virt(page_to_phys(page));
 		reset_ctx(head);
 		ctx = &head->ctx;
-		frm = &head->frm;
+		frm = head->frame;
 		xdp->frame_cnt++;
 
 		act = bpf_prog_run_xdp(prog, ctx);
diff --git a/tools/testing/selftests/bpf/prog_tests/xdp_do_redirect.c b/tools/testing/selftests/bpf/prog_tests/xdp_do_redirect.c
index 2666c84dbd01..7271a18ab3e2 100644
--- a/tools/testing/selftests/bpf/prog_tests/xdp_do_redirect.c
+++ b/tools/testing/selftests/bpf/prog_tests/xdp_do_redirect.c
@@ -65,12 +65,13 @@  static int attach_tc_prog(struct bpf_tc_hook *hook, int fd)
 }
 
 /* The maximum permissible size is: PAGE_SIZE - sizeof(struct xdp_page_head) -
- * sizeof(struct skb_shared_info) - XDP_PACKET_HEADROOM = 3368 bytes
+ * SKB_DATA_ALIGN(sizeof(struct skb_shared_info)) - XDP_PACKET_HEADROOM =
+ * 3408 bytes for 64-byte cacheline and 3216 for 256-byte one.
  */
 #if defined(__s390x__)
-#define MAX_PKT_SIZE 3176
+#define MAX_PKT_SIZE 3216
 #else
-#define MAX_PKT_SIZE 3368
+#define MAX_PKT_SIZE 3408
 #endif
 static void test_max_pkt_size(int fd)
 {