| Message ID | 20240409140914.4105429-1-edumazet@google.com (mailing list archive) |
|---|---|
| State | Accepted |
| Commit | 9b9fd45869e744bea7d32a94793736e3d9fd7d26 |
| Delegated to: | Netdev Maintainers |
| Headers | show |
| Series | [net-next] tcp: tweak tcp_sock_write_txrx size assertion | expand |
On Tue, Apr 09, 2024 at 02:09:14PM +0000, Eric Dumazet wrote: > I forgot 32bit arches might have 64bit alignment for u64 > fields. > > tcp_sock_write_txrx group does not contain pointers, > but two u64 fields. It is possible that on 32bit kernel, > a 32bit hole is before tp->tcp_clock_cache. > > I will try to remember a group can be bigger on 32bit > kernels in the future. > > With help from Vladimir Oltean. > > Fixes: d2c3a7eb1afa ("tcp: more struct tcp_sock adjustments") > Reported-by: kernel test robot <lkp@intel.com> > Closes: https://lore.kernel.org/oe-kbuild-all/202404082207.HCEdQhUO-lkp@intel.com/ > Signed-off-by: Eric Dumazet <edumazet@google.com> > Cc: Vladimir Oltean <vladimir.oltean@nxp.com> > --- Tested-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Hello: This patch was applied to netdev/net-next.git (main) by Jakub Kicinski <kuba@kernel.org>: On Tue, 9 Apr 2024 14:09:14 +0000 you wrote: > I forgot 32bit arches might have 64bit alignment for u64 > fields. > > tcp_sock_write_txrx group does not contain pointers, > but two u64 fields. It is possible that on 32bit kernel, > a 32bit hole is before tp->tcp_clock_cache. > > [...] Here is the summary with links: - [net-next] tcp: tweak tcp_sock_write_txrx size assertion https://git.kernel.org/netdev/net-next/c/9b9fd45869e7 You are awesome, thank you!
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c index b07aa71b24ec147d2f1f148e288231b3a492fb5a..e1bf468e0d22aad01532dce6e226ca095d578b91 100644 --- a/net/ipv4/tcp.c +++ b/net/ipv4/tcp.c @@ -4673,7 +4673,11 @@ static void __init tcp_struct_check(void) CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_txrx, app_limited); CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_txrx, rcv_wnd); CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_txrx, rx_opt); - CACHELINE_ASSERT_GROUP_SIZE(struct tcp_sock, tcp_sock_write_txrx, 92); + + /* 32bit arches with 8byte alignment on u64 fields might need padding + * before tcp_clock_cache. + */ + CACHELINE_ASSERT_GROUP_SIZE(struct tcp_sock, tcp_sock_write_txrx, 92 + 4); /* RX read-write hotpath cache lines */ CACHELINE_ASSERT_GROUP_MEMBER(struct tcp_sock, tcp_sock_write_rx, bytes_received);
I forgot 32bit arches might have 64bit alignment for u64 fields. tcp_sock_write_txrx group does not contain pointers, but two u64 fields. It is possible that on 32bit kernel, a 32bit hole is before tp->tcp_clock_cache. I will try to remember a group can be bigger on 32bit kernels in the future. With help from Vladimir Oltean. Fixes: d2c3a7eb1afa ("tcp: more struct tcp_sock adjustments") Reported-by: kernel test robot <lkp@intel.com> Closes: https://lore.kernel.org/oe-kbuild-all/202404082207.HCEdQhUO-lkp@intel.com/ Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Vladimir Oltean <vladimir.oltean@nxp.com> --- net/ipv4/tcp.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-)