| Message ID | 20240428111640.27306-1-linyunsheng@huawei.com (mailing list archive) |
|---|---|
| State | Accepted |
| Commit | 9f8eeea1643c213c0e1ad2e546a15536200d216b |
| Delegated to: | Netdev Maintainers |
| Headers | show |
| Series | [v2] rxrpc: Fix using alignmask being zero for __page_frag_alloc_align() | expand |
On Sun, 2024-04-28 at 19:16 +0800, Yunsheng Lin wrote: > rxrpc_alloc_data_txbuf() may be called with data_align being > zero in none_alloc_txbuf() and rxkad_alloc_txbuf(), data_align > is supposed to be an order-based alignment value, but zero is > not a valid order-based alignment value, and '~(data_align - 1)' > doesn't result in a valid mask-based alignment value for > __page_frag_alloc_align(). > > Fix it by passing a valid order-based alignment value in > none_alloc_txbuf() and rxkad_alloc_txbuf(). > > Also use page_frag_alloc_align() expecting an order-based > alignment value in rxrpc_alloc_data_txbuf() to avoid doing the > alignment converting operation and to catch possible invalid > alignment value in the future. Remove the 'if (data_align)' > checking too, as it is always true for a valid order-based > alignment value. > > Fixes: 6b2536462fd4 ("rxrpc: Fix use of changed alignment param to page_frag_alloc_align()") > Fixes: 49489bb03a50 ("rxrpc: Do zerocopy using MSG_SPLICE_PAGES and page frags") > CC: David Howells <dhowells@redhat.com> > CC: Alexander Duyck <alexander.duyck@gmail.com> > Signed-off-by: Yunsheng Lin <linyunsheng@huawei.com> > Acked-by: David Howells <dhowells@redhat.com> Double checking before doing dumb mistakes... @David: I assume from your ack you are fine with this patch being merged via the net tree, am I correct? Thanks, Paolo
Paolo Abeni <pabeni@redhat.com> wrote: > @David: I assume from your ack you are fine with this patch being > merged via the net tree, am I correct? Yes, the net tree is fine. David
Hello: This patch was applied to netdev/net.git (main) by Jakub Kicinski <kuba@kernel.org>: On Sun, 28 Apr 2024 19:16:38 +0800 you wrote: > rxrpc_alloc_data_txbuf() may be called with data_align being > zero in none_alloc_txbuf() and rxkad_alloc_txbuf(), data_align > is supposed to be an order-based alignment value, but zero is > not a valid order-based alignment value, and '~(data_align - 1)' > doesn't result in a valid mask-based alignment value for > __page_frag_alloc_align(). > > [...] Here is the summary with links: - [v2] rxrpc: Fix using alignmask being zero for __page_frag_alloc_align() https://git.kernel.org/netdev/net/c/9f8eeea1643c You are awesome, thank you!
diff --git a/net/rxrpc/insecure.c b/net/rxrpc/insecure.c index f2701068ed9e..6716c021a532 100644 --- a/net/rxrpc/insecure.c +++ b/net/rxrpc/insecure.c @@ -19,7 +19,7 @@ static int none_init_connection_security(struct rxrpc_connection *conn, */ static struct rxrpc_txbuf *none_alloc_txbuf(struct rxrpc_call *call, size_t remain, gfp_t gfp) { - return rxrpc_alloc_data_txbuf(call, min_t(size_t, remain, RXRPC_JUMBO_DATALEN), 0, gfp); + return rxrpc_alloc_data_txbuf(call, min_t(size_t, remain, RXRPC_JUMBO_DATALEN), 1, gfp); } static int none_secure_packet(struct rxrpc_call *call, struct rxrpc_txbuf *txb) diff --git a/net/rxrpc/rxkad.c b/net/rxrpc/rxkad.c index f1a68270862d..48a1475e6b06 100644 --- a/net/rxrpc/rxkad.c +++ b/net/rxrpc/rxkad.c @@ -155,7 +155,7 @@ static struct rxrpc_txbuf *rxkad_alloc_txbuf(struct rxrpc_call *call, size_t rem switch (call->conn->security_level) { default: space = min_t(size_t, remain, RXRPC_JUMBO_DATALEN); - return rxrpc_alloc_data_txbuf(call, space, 0, gfp); + return rxrpc_alloc_data_txbuf(call, space, 1, gfp); case RXRPC_SECURITY_AUTH: shdr = sizeof(struct rxkad_level1_hdr); break; diff --git a/net/rxrpc/txbuf.c b/net/rxrpc/txbuf.c index e0679658d9de..c3913d8a50d3 100644 --- a/net/rxrpc/txbuf.c +++ b/net/rxrpc/txbuf.c @@ -21,20 +21,20 @@ struct rxrpc_txbuf *rxrpc_alloc_data_txbuf(struct rxrpc_call *call, size_t data_ { struct rxrpc_wire_header *whdr; struct rxrpc_txbuf *txb; - size_t total, hoff = 0; + size_t total, hoff; void *buf; txb = kmalloc(sizeof(*txb), gfp); if (!txb) return NULL; - if (data_align) - hoff = round_up(sizeof(*whdr), data_align) - sizeof(*whdr); + hoff = round_up(sizeof(*whdr), data_align) - sizeof(*whdr); total = hoff + sizeof(*whdr) + data_size; + data_align = umax(data_align, L1_CACHE_BYTES); mutex_lock(&call->conn->tx_data_alloc_lock); - buf = __page_frag_alloc_align(&call->conn->tx_data_alloc, total, gfp, - ~(data_align - 1) & ~(L1_CACHE_BYTES - 1)); + buf = page_frag_alloc_align(&call->conn->tx_data_alloc, total, gfp, + data_align); mutex_unlock(&call->conn->tx_data_alloc_lock); if (!buf) { kfree(txb);
rxrpc_alloc_data_txbuf() may be called with data_align being zero in none_alloc_txbuf() and rxkad_alloc_txbuf(), data_align is supposed to be an order-based alignment value, but zero is not a valid order-based alignment value, and '~(data_align - 1)' doesn't result in a valid mask-based alignment value for __page_frag_alloc_align(). Fix it by passing a valid order-based alignment value in none_alloc_txbuf() and rxkad_alloc_txbuf(). Also use page_frag_alloc_align() expecting an order-based alignment value in rxrpc_alloc_data_txbuf() to avoid doing the alignment converting operation and to catch possible invalid alignment value in the future. Remove the 'if (data_align)' checking too, as it is always true for a valid order-based alignment value. Fixes: 6b2536462fd4 ("rxrpc: Fix use of changed alignment param to page_frag_alloc_align()") Fixes: 49489bb03a50 ("rxrpc: Do zerocopy using MSG_SPLICE_PAGES and page frags") CC: David Howells <dhowells@redhat.com> CC: Alexander Duyck <alexander.duyck@gmail.com> Signed-off-by: Yunsheng Lin <linyunsheng@huawei.com> Acked-by: David Howells <dhowells@redhat.com> --- V2: use 'umax' and drop 'U' as suggested by David. --- net/rxrpc/insecure.c | 2 +- net/rxrpc/rxkad.c | 2 +- net/rxrpc/txbuf.c | 10 +++++----- 3 files changed, 7 insertions(+), 7 deletions(-)