On Tue, 19 Jan 2021 16:42:14 +0100 Rasmus Villemoes wrote: > On 28/12/2020 23.24, Jakub Kicinski wrote: > > On Wed, 23 Dec 2020 15:45:32 +0100 Rasmus Villemoes wrote: > >> Wireshark says that the MRP test packets cannot be decoded - and the > >> reason for that is that there's a two-byte hole filled with garbage > >> between the "transitions" and "timestamp" members. > >> > >> So Wireshark decodes the two garbage bytes and the top two bytes of > >> the timestamp written by the kernel as the timestamp value (which thus > >> fluctuates wildly), and interprets the lower two bytes of the > >> timestamp as a new (type, length) pair, which is of course broken. > >> > >> While my copy of the MRP standard is still under way [*], I cannot > >> imagine the standard specifying a two-byte hole here, and whoever > >> wrote the Wireshark decoding code seems to agree with that. > >> > >> The struct definitions live under include/uapi/, but they are not > >> really part of any kernel<->userspace API/ABI, so fixing the > >> definitions by adding the packed attribute should not cause any > >> compatibility issues. > >> > >> The remaining on-the-wire packet formats likely also don't contain > >> holes, but pahole and manual inspection says the current definitions > >> suffice. So adding the packed attribute to those is not strictly > >> needed, but might be done for good measure. > >> > >> [*] I will never understand how something hidden behind a +1000$ > >> paywall can be called a standard. > >> > >> Signed-off-by: Rasmus Villemoes <rasmus.villem...@prevas.dk> > >> --- > >> include/uapi/linux/mrp_bridge.h | 4 ++-- > >> 1 file changed, 2 insertions(+), 2 deletions(-) > >> > >> diff --git a/include/uapi/linux/mrp_bridge.h > >> b/include/uapi/linux/mrp_bridge.h > >> index 6aeb13ef0b1e..d1d0cf65916d 100644 > >> --- a/include/uapi/linux/mrp_bridge.h > >> +++ b/include/uapi/linux/mrp_bridge.h > >> @@ -96,7 +96,7 @@ struct br_mrp_ring_test_hdr { > >> __be16 state; > >> __be16 transitions; > >> __be32 timestamp; > >> -}; > >> +} __attribute__((__packed__)); > >> > >> struct br_mrp_ring_topo_hdr { > >> __be16 prio; > >> @@ -141,7 +141,7 @@ struct br_mrp_in_test_hdr { > >> __be16 state; > >> __be16 transitions; > >> __be32 timestamp; > >> -}; > >> +} __attribute__((__packed__)); > >> > >> struct br_mrp_in_topo_hdr { > >> __u8 sa[ETH_ALEN]; > > > > Can we use this opportunity to move the definitions of these structures > > out of the uAPI to a normal kernel header? > > Jakub, can we apply this patch to net, then later move the definitions > out of uapi (and deleting the unused ones in the process)?
This has been lost in the patchwork archives already, we'll need a fresh posting anyway. Why not do the move as a part of the same series? Doesn't seems like much work, unless I'm missing something..