The test_run code detects whether a page has been modified and
re-initialises the xdp_frame structure if it has, using
xdp_update_frame_from_buff(). However, xdp_update_frame_from_buff()
doesn't touch frame->mem, so that wasn't correctly re-initialised, which
led to the pages from page_pool not being returned correctly. Syzbot
noticed this as a memory leak.
Fix this by also copying the frame->mem structure when re-initialising
the frame, like we do on initialisation of a new page from page_pool.
Reported-by: syzbot+d121e098da06af416d23@xxxxxxxxxxxxxxxxxxxxxxxxx
Tested-by: syzbot+d121e098da06af416d23@xxxxxxxxxxxxxxxxxxxxxxxxx
Fixes: e5995bc7e2ba ("bpf, test_run: fix crashes due to XDP frame overwriting/corruption")
Fixes: b530e9e1063e ("bpf: Add "live packet" mode for XDP in BPF_PROG_RUN")
Signed-off-by: Toke Høiland-Jørgensen <toke@xxxxxxxxxx>
---
net/bpf/test_run.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/net/bpf/test_run.c b/net/bpf/test_run.c
index 6d7a442ceb89be15501069655a51671d6ddfaf0e..501ec4249fedc3d34fe39aff50eea66f82b88a11 100644
--- a/net/bpf/test_run.c
+++ b/net/bpf/test_run.c
@@ -246,6 +246,7 @@ static void reset_ctx(struct xdp_page_head *head)
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->frame);
+ head->frame->mem = head->orig_ctx.rxq->mem;
}
static int xdp_recv_frames(struct xdp_frame **frames, int nframes,
---
base-commit: d0b98f6a17a5cb336121302bce0c97eb5fe32d16
change-id: 20241030-test-run-mem-fix-ac835c75ec3f
Best regards,
--
Toke Høiland-Jørgensen <toke@xxxxxxxxxx>
