summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* rxrpc: Fix NULL pointer deref due to call->conn being cleared on disconnectDavid Howells2020-02-035-24/+15
| | | | | | | | | | | | | | | | When a call is disconnected, the connection pointer from the call is cleared to make sure it isn't used again and to prevent further attempted transmission for the call. Unfortunately, there might be a daemon trying to use it at the same time to transmit a packet. Fix this by keeping call->conn set, but setting a flag on the call to indicate disconnection instead. Remove also the bits in the transmission functions where the conn pointer is checked and a ref taken under spinlock as this is now redundant. Fixes: 8d94aa381dab ("rxrpc: Calls shouldn't hold socket refs") Signed-off-by: David Howells <dhowells@redhat.com>
* rxrpc: Fix missing active use pinning of rxrpc_local objectDavid Howells2020-01-305-40/+62
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The introduction of a split between the reference count on rxrpc_local objects and the usage count didn't quite go far enough. A number of kernel work items need to make use of the socket to perform transmission. These also need to get an active count on the local object to prevent the socket from being closed. Fix this by getting the active count in those places. Also split out the raw active count get/put functions as these places tend to hold refs on the rxrpc_local object already, so getting and putting an extra object ref is just a waste of time. The problem can lead to symptoms like: BUG: kernel NULL pointer dereference, address: 0000000000000018 .. CPU: 2 PID: 818 Comm: kworker/u9:0 Not tainted 5.5.0-fscache+ #51 ... RIP: 0010:selinux_socket_sendmsg+0x5/0x13 ... Call Trace: security_socket_sendmsg+0x2c/0x3e sock_sendmsg+0x1a/0x46 rxrpc_send_keepalive+0x131/0x1ae rxrpc_peer_keepalive_worker+0x219/0x34b process_one_work+0x18e/0x271 worker_thread+0x1a3/0x247 kthread+0xe6/0xeb ret_from_fork+0x1f/0x30 Fixes: 730c5fd42c1e ("rxrpc: Fix local endpoint refcounting") Signed-off-by: David Howells <dhowells@redhat.com>
* rxrpc: Fix insufficient receive notification generationDavid Howells2020-01-301-4/+2
| | | | | | | | | | | | | | | | | In rxrpc_input_data(), rxrpc_notify_socket() is called if the base sequence number of the packet is immediately following the hard-ack point at the end of the function. However, this isn't sufficient, since the recvmsg side may have been advancing the window and then overrun the position in which we're adding - at which point rx_hard_ack >= seq0 and no notification is generated. Fix this by always generating a notification at the end of the input function. Without this, a long call may stall, possibly indefinitely. Fixes: 248f219cb8bc ("rxrpc: Rewrite the data and ack handling code") Signed-off-by: David Howells <dhowells@redhat.com>
* rxrpc: Fix use-after-free in rxrpc_put_local()David Howells2020-01-301-1/+4
| | | | | | | | | Fix rxrpc_put_local() to not access local->debug_id after calling atomic_dec_return() as, unless that returned n==0, we no longer have the right to access the object. Fixes: 06d9532fa6b3 ("rxrpc: Fix read-after-free in rxrpc_queue_local()") Signed-off-by: David Howells <dhowells@redhat.com>
* net/core: Do not clear VF index for node/port GUIDs queryLeon Romanovsky2020-01-301-2/+2
| | | | | | | | | | | VF numbers were assigned to node_guid and port_guid, but cleared right before such query calls were issued. It caused to return node/port GUIDs of VF index 0 for all VFs. Fixes: 30aad41721e0 ("net/core: Add support for getting VF GUIDs") Reported-by: Adrian Chiris <adrianc@mellanox.com> Signed-off-by: Leon Romanovsky <leonro@mellanox.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* mptcp: Fix undefined mptcp_handle_ipv6_mapped for modular IPV6Geert Uytterhoeven2020-01-303-12/+9
| | | | | | | | | | | | | | | | | | | | If CONFIG_MPTCP=y, CONFIG_MPTCP_IPV6=n, and CONFIG_IPV6=m: ERROR: "mptcp_handle_ipv6_mapped" [net/ipv6/ipv6.ko] undefined! This does not happen if CONFIG_MPTCP_IPV6=y, as CONFIG_MPTCP_IPV6 selects CONFIG_IPV6, and thus forces CONFIG_IPV6 builtin. As exporting a symbol for an empty function would be a bit wasteful, fix this by providing a dummy version of mptcp_handle_ipv6_mapped() for the CONFIG_MPTCP_IPV6=n case. Rename mptcp_handle_ipv6_mapped() to mptcpv6_handle_mapped(), to make it clear this is a pure-IPV6 function, just like mptcpv6_init(). Fixes: cec37a6e41aae7bf ("mptcp: Handle MP_CAPABLE options for outgoing connections") Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org> Signed-off-by: David S. Miller <davem@davemloft.net>
* net: drop_monitor: Use kstrdupJoe Perches2020-01-301-6/+2
| | | | | | | | | Convert the equivalent but rather odd uses of kmemdup with __GFP_ZERO to the more common kstrdup and avoid unnecessary zeroing of copied over memory. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* udp: document udp_rcv_segment special case for looped packetsWillem de Bruijn2020-01-301-0/+7
| | | | | | | | | | | | | | | | Commit 6cd021a58c18a ("udp: segment looped gso packets correctly") fixes an issue with rare udp gso multicast packets looped onto the receive path. The stable backport makes the narrowest change to target only these packets, when needed. As opposed to, say, expanding __udp_gso_segment, which is harder to reason to be free from unintended side-effects. But the resulting code is hardly self-describing. Document its purpose and rationale. Signed-off-by: Willem de Bruijn <willemb@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* mptcp: MPTCP_HMAC_TEST should depend on MPTCPGeert Uytterhoeven2020-01-301-2/+4
| | | | | | | | | | | | | | | As the MPTCP HMAC test is integrated into the MPTCP code, it can be built only when MPTCP is enabled. Hence when MPTCP is disabled, asking the user if the test code should be enabled is futile. Wrap the whole block of MPTCP-specific config options inside a check for MPTCP. While at it, drop the "default n" for MPTCP_HMAC_TEST, as that is the default anyway. Fixes: 65492c5a6ab5df50 ("mptcp: move from sha1 (v0) to sha256 (v1)") Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org> Reviewed-by: Mat Martineau <mathew.j.martineau@linux.intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* mptcp: Fix incorrect IPV6 dependency checkGeert Uytterhoeven2020-01-301-1/+1
| | | | | | | | | | | | | | | If CONFIG_MPTCP=y, CONFIG_MPTCP_IPV6=n, and CONFIG_IPV6=m: net/mptcp/protocol.o: In function `__mptcp_tcp_fallback': protocol.c:(.text+0x786): undefined reference to `inet6_stream_ops' Fix this by checking for CONFIG_MPTCP_IPV6 instead of CONFIG_IPV6, like is done in all other places in the mptcp code. Fixes: 8ab183deb26a3b79 ("mptcp: cope with later TCP fallback") Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org> Reviewed-by: Mat Martineau <mathew.j.martineau@linux.intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* Revert "MAINTAINERS: mptcp@ mailing list is moderated"Mat Martineau2020-01-291-1/+1
| | | | | | | | | | | | | This reverts commit 74759e1693311a8d1441de836c4080c192374238. mptcp@lists.01.org accepts messages from non-subscribers. There was an invisible and unexpected server-wide rule limiting the number of recipients for subscribers and non-subscribers alike, and that has now been turned off for this list. Cc: Randy Dunlap <rdunlap@infradead.org> Signed-off-by: Mat Martineau <mathew.j.martineau@linux.intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* Merge branch 'mptcp-fix-sockopt-crash-and-lockdep-splats'David S. Miller2020-01-296-25/+46
|\ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Florian Westphal says: ==================== mptcp: fix sockopt crash and lockdep splats Christoph Paasch reported a few bugs and lockdep splats triggered by syzkaller. One patch fixes a crash in set/getsockopt. Two patches fix lockdep splats related to the order in which RTNL and socket lock are taken. Last patch fixes out-of-bounds access when TCP syncookies are used. Change since last iteration on mptcp-list: - add needed refcount in patch 2 - call tcp_get/setsockopt directly in patch 2 Other patches unchanged except minor amends to commit messages. ==================== Signed-off-by: David S. Miller <davem@davemloft.net>
| * mptcp: handle tcp fallback when using syn cookiesFlorian Westphal2020-01-295-3/+14
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | We can't deal with syncookie mode yet, the syncookie rx path will create tcp reqsk, i.e. we get OOB access because we treat tcp reqsk as mptcp reqsk one: TCP: SYN flooding on port 20002. Sending cookies. BUG: KASAN: slab-out-of-bounds in subflow_syn_recv_sock+0x451/0x4d0 net/mptcp/subflow.c:191 Read of size 1 at addr ffff8881167bc148 by task syz-executor099/2120 subflow_syn_recv_sock+0x451/0x4d0 net/mptcp/subflow.c:191 tcp_get_cookie_sock+0xcf/0x520 net/ipv4/syncookies.c:209 cookie_v6_check+0x15a5/0x1e90 net/ipv6/syncookies.c:252 tcp_v6_cookie_check net/ipv6/tcp_ipv6.c:1123 [inline] [..] Bug can be reproduced via "sysctl net.ipv4.tcp_syncookies=2". Note that MPTCP should work with syncookies (4th ack would carry needed state), but it appears better to sort that out in -next so do tcp fallback for now. I removed the MPTCP ifdef for tcp_rsk "is_mptcp" member because if (IS_ENABLED()) is easier to read than "#ifdef IS_ENABLED()/#endif" pair. Cc: Eric Dumazet <edumazet@google.com> Fixes: cec37a6e41aae7bf ("mptcp: Handle MP_CAPABLE options for outgoing connections") Reported-by: Christoph Paasch <cpaasch@apple.com> Tested-by: Christoph Paasch <cpaasch@apple.com> Signed-off-by: Florian Westphal <fw@strlen.de> Signed-off-by: David S. Miller <davem@davemloft.net>
| * mptcp: avoid a lockdep splat when mcast group was joinedFlorian Westphal2020-01-291-2/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | syzbot triggered following lockdep splat: ffffffff82d2cd40 (rtnl_mutex){+.+.}, at: ip_mc_drop_socket+0x52/0x180 but task is already holding lock: ffff8881187a2310 (sk_lock-AF_INET){+.+.}, at: mptcp_close+0x18/0x30 which lock already depends on the new lock. the existing dependency chain (in reverse order) is: -> #1 (sk_lock-AF_INET){+.+.}: lock_acquire+0xee/0x230 lock_sock_nested+0x89/0xc0 do_ip_setsockopt.isra.0+0x335/0x22f0 ip_setsockopt+0x35/0x60 tcp_setsockopt+0x5d/0x90 __sys_setsockopt+0xf3/0x190 __x64_sys_setsockopt+0x61/0x70 do_syscall_64+0x72/0x300 entry_SYSCALL_64_after_hwframe+0x49/0xbe -> #0 (rtnl_mutex){+.+.}: check_prevs_add+0x2b7/0x1210 __lock_acquire+0x10b6/0x1400 lock_acquire+0xee/0x230 __mutex_lock+0x120/0xc70 ip_mc_drop_socket+0x52/0x180 inet_release+0x36/0xe0 __sock_release+0xfd/0x130 __mptcp_close+0xa8/0x1f0 inet_release+0x7f/0xe0 __sock_release+0x69/0x130 sock_close+0x18/0x20 __fput+0x179/0x400 task_work_run+0xd5/0x110 do_exit+0x685/0x1510 do_group_exit+0x7e/0x170 __x64_sys_exit_group+0x28/0x30 do_syscall_64+0x72/0x300 entry_SYSCALL_64_after_hwframe+0x49/0xbe The trigger is: socket(AF_INET, SOCK_STREAM, 0x106 /* IPPROTO_MPTCP */) = 4 setsockopt(4, SOL_IP, MCAST_JOIN_GROUP, {gr_interface=7, gr_group={sa_family=AF_INET, sin_port=htons(20003), sin_addr=inet_addr("224.0.0.2")}}, 136) = 0 exit(0) Which results in a call to rtnl_lock while we are holding the parent mptcp socket lock via mptcp_close -> lock_sock(msk) -> inet_release -> ip_mc_drop_socket -> rtnl_lock(). >From lockdep point of view we thus have both 'rtnl_lock; lock_sock' and 'lock_sock; rtnl_lock'. Fix this by stealing the msk conn_list and doing the subflow close without holding the msk lock. Fixes: cec37a6e41aae7bf ("mptcp: Handle MP_CAPABLE options for outgoing connections") Reported-by: Christoph Paasch <cpaasch@apple.com> Signed-off-by: Florian Westphal <fw@strlen.de> Signed-off-by: David S. Miller <davem@davemloft.net>
| * mptcp: fix panic on user pointer accessFlorian Westphal2020-01-291-18/+22
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Its not possible to call the kernel_(s|g)etsockopt functions here, the address points to user memory: General protection fault in user access. Non-canonical address? WARNING: CPU: 1 PID: 5352 at arch/x86/mm/extable.c:77 ex_handler_uaccess+0xba/0xe0 arch/x86/mm/extable.c:77 Kernel panic - not syncing: panic_on_warn set ... [..] Call Trace: fixup_exception+0x9d/0xcd arch/x86/mm/extable.c:178 general_protection+0x2d/0x40 arch/x86/entry/entry_64.S:1202 do_ip_getsockopt+0x1f6/0x1860 net/ipv4/ip_sockglue.c:1323 ip_getsockopt+0x87/0x1c0 net/ipv4/ip_sockglue.c:1561 tcp_getsockopt net/ipv4/tcp.c:3691 [inline] tcp_getsockopt+0x8c/0xd0 net/ipv4/tcp.c:3685 kernel_getsockopt+0x121/0x1f0 net/socket.c:3736 mptcp_getsockopt+0x69/0x90 net/mptcp/protocol.c:830 __sys_getsockopt+0x13a/0x220 net/socket.c:2175 We can call tcp_get/setsockopt functions instead. Doing so fixes crashing, but still leaves rtnl related lockdep splat: WARNING: possible circular locking dependency detected 5.5.0-rc6 #2 Not tainted ------------------------------------------------------ syz-executor.0/16334 is trying to acquire lock: ffffffff84f7a080 (rtnl_mutex){+.+.}, at: do_ip_setsockopt.isra.0+0x277/0x3820 net/ipv4/ip_sockglue.c:644 but task is already holding lock: ffff888116503b90 (sk_lock-AF_INET){+.+.}, at: lock_sock include/net/sock.h:1516 [inline] ffff888116503b90 (sk_lock-AF_INET){+.+.}, at: mptcp_setsockopt+0x28/0x90 net/mptcp/protocol.c:1284 which lock already depends on the new lock. the existing dependency chain (in reverse order) is: -> #1 (sk_lock-AF_INET){+.+.}: lock_sock_nested+0xca/0x120 net/core/sock.c:2944 lock_sock include/net/sock.h:1516 [inline] do_ip_setsockopt.isra.0+0x281/0x3820 net/ipv4/ip_sockglue.c:645 ip_setsockopt+0x44/0xf0 net/ipv4/ip_sockglue.c:1248 udp_setsockopt+0x5d/0xa0 net/ipv4/udp.c:2639 __sys_setsockopt+0x152/0x240 net/socket.c:2130 __do_sys_setsockopt net/socket.c:2146 [inline] __se_sys_setsockopt net/socket.c:2143 [inline] __x64_sys_setsockopt+0xba/0x150 net/socket.c:2143 do_syscall_64+0xbd/0x5b0 arch/x86/entry/common.c:294 entry_SYSCALL_64_after_hwframe+0x49/0xbe -> #0 (rtnl_mutex){+.+.}: check_prev_add kernel/locking/lockdep.c:2475 [inline] check_prevs_add kernel/locking/lockdep.c:2580 [inline] validate_chain kernel/locking/lockdep.c:2970 [inline] __lock_acquire+0x1fb2/0x4680 kernel/locking/lockdep.c:3954 lock_acquire+0x127/0x330 kernel/locking/lockdep.c:4484 __mutex_lock_common kernel/locking/mutex.c:956 [inline] __mutex_lock+0x158/0x1340 kernel/locking/mutex.c:1103 do_ip_setsockopt.isra.0+0x277/0x3820 net/ipv4/ip_sockglue.c:644 ip_setsockopt+0x44/0xf0 net/ipv4/ip_sockglue.c:1248 tcp_setsockopt net/ipv4/tcp.c:3159 [inline] tcp_setsockopt+0x8c/0xd0 net/ipv4/tcp.c:3153 kernel_setsockopt+0x121/0x1f0 net/socket.c:3767 mptcp_setsockopt+0x69/0x90 net/mptcp/protocol.c:1288 __sys_setsockopt+0x152/0x240 net/socket.c:2130 __do_sys_setsockopt net/socket.c:2146 [inline] __se_sys_setsockopt net/socket.c:2143 [inline] __x64_sys_setsockopt+0xba/0x150 net/socket.c:2143 do_syscall_64+0xbd/0x5b0 arch/x86/entry/common.c:294 entry_SYSCALL_64_after_hwframe+0x49/0xbe other info that might help us debug this: Possible unsafe locking scenario: CPU0 CPU1 ---- ---- lock(sk_lock-AF_INET); lock(rtnl_mutex); lock(sk_lock-AF_INET); lock(rtnl_mutex); The lockdep complaint is because we hold mptcp socket lock when calling the sk_prot get/setsockopt handler, and those might need to acquire the rtnl mutex. Normally, order is: rtnl_lock(sk) -> lock_sock Whereas for mptcp the order is lock_sock(mptcp_sk) rtnl_lock -> lock_sock(subflow_sk) We can avoid this by releasing the mptcp socket lock early, but, as Paolo points out, we need to get/put the subflow socket refcount before doing so to avoid race with concurrent close(). Fixes: 717e79c867ca5 ("mptcp: Add setsockopt()/getsockopt() socket operations") Reported-by: Christoph Paasch <cpaasch@apple.com> Signed-off-by: Florian Westphal <fw@strlen.de> Signed-off-by: David S. Miller <davem@davemloft.net>
| * mptcp: defer freeing of cached ext until last momentFlorian Westphal2020-01-291-2/+4
|/ | | | | | | | | | | | | access to msk->cached_ext is only legal if the msk is locked or all concurrent accesses are impossible. Furthermore, once we start to tear down, we must make sure nothing else can step in and allocate a new cached ext. So place this code in the destroy callback where it belongs. Signed-off-by: Florian Westphal <fw@strlen.de> Signed-off-by: David S. Miller <davem@davemloft.net>
* net: mvneta: fix XDP support if sw bm is used as fallbackLorenzo Bianconi2020-01-291-3/+7
| | | | | | | | | | | | | | In order to fix XDP support if sw buffer management is used as fallback for hw bm devices, define MVNETA_SKB_HEADROOM as maximum between XDP_PACKET_HEADROOM and NET_SKB_PAD and let the hw aligns the IP header to 4-byte boundary. Fix rx_offset_correction initialization if mvneta_bm_port_init fails in mvneta_resume routine Fixes: 0db51da7a8e9 ("net: mvneta: add basic XDP support") Tested-by: Sven Auhagen <sven.auhagen@voleatech.de> Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: David S. Miller <davem@davemloft.net>
* sch_choke: Use kvcallocJoe Perches2020-01-291-1/+1
| | | | | | | | Convert the use of kvmalloc_array with __GFP_ZERO to the equivalent kvcalloc. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* mptcp: Fix build with PROC_FS disabled.David S. Miller2020-01-291-0/+2
| | | | | | | | net/mptcp/subflow.c: In function ‘mptcp_subflow_create_socket’: net/mptcp/subflow.c:624:25: error: ‘struct netns_core’ has no member named ‘sock_inuse’ Reported-by: Randy Dunlap <rdunlap@infradead.org> Signed-off-by: David S. Miller <davem@davemloft.net>
* MAINTAINERS: mptcp@ mailing list is moderatedRandy Dunlap2020-01-291-1/+1
| | | | | | | | | | | | Note that mptcp@lists.01.org is moderated, like we note for other mailing lists. Signed-off-by: Randy Dunlap <rdunlap@infradead.org> Cc: Mat Martineau <mathew.j.martineau@linux.intel.com> Cc: Matthieu Baerts <matthieu.baerts@tessares.net> Cc: netdev@vger.kernel.org Cc: mptcp@lists.01.org Signed-off-by: David S. Miller <davem@davemloft.net>
* Merge branch 'for-v5.6' of ↵Linus Torvalds2020-01-291-1/+0
|\ | | | | | | | | | | | | | | | | | | git://git.kernel.org:/pub/scm/linux/kernel/git/jmorris/linux-security Pull security subsystem update from James Morris: "Just one minor fix this time" * 'for-v5.6' of git://git.kernel.org:/pub/scm/linux/kernel/git/jmorris/linux-security: security: remove EARLY_LSM_COUNT which never used
| * security: remove EARLY_LSM_COUNT which never usedAlex Shi2020-01-271-1/+0
| | | | | | | | | | | | | | | | | | This macro is never used from it was introduced in commit e6b1db98cf4d5 ("security: Support early LSMs"), better to remove it. Signed-off-by: Alex Shi <alex.shi@linux.alibaba.com> Acked-by: Serge Hallyn <serge@hallyn.com> Signed-off-by: James Morris <jmorris@namei.org>
* | Merge branch 'next-integrity' of ↵Linus Torvalds2020-01-2914-40/+540
|\ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | git://git.kernel.org/pub/scm/linux/kernel/git/zohar/linux-integrity Pull IMA updates from Mimi Zohar: "Two new features - measuring certificates and querying IMA for a file hash - and three bug fixes: - Measuring certificates is like the rest of IMA, based on policy, but requires loading a custom policy. Certificates loaded onto a keyring, for example during early boot, before a custom policy has been loaded, are queued and only processed after loading the custom policy. - IMA calculates and caches files hashes. Other kernel subsystems, and possibly kernel modules, are interested in accessing these cached file hashes. The bug fixes prevent classifying a file short read (e.g. shutdown) as an invalid file signature, add a missing blank when displaying the securityfs policy rules containing LSM labels, and, lastly, fix the handling of the IMA policy information for unknown LSM labels" * 'next-integrity' of git://git.kernel.org/pub/scm/linux/kernel/git/zohar/linux-integrity: IMA: Defined delayed workqueue to free the queued keys IMA: Call workqueue functions to measure queued keys IMA: Define workqueue for early boot key measurements IMA: pre-allocate buffer to hold keyrings string ima: ima/lsm policy rule loading logic bug fixes ima: add the ability to query the cached hash of a given file ima: Add a space after printing LSM rules for readability IMA: fix measuring asymmetric keys Kconfig IMA: Read keyrings= option from the IMA policy IMA: Add support to limit measuring keys KEYS: Call the IMA hook to measure keys IMA: Define an IMA hook to measure keys IMA: Add KEY_CHECK func to measure keys IMA: Check IMA policy flag ima: avoid appraise error for hash calc interrupt
| * \ Merge branch 'next-integrity.defer-measuring-keys' into next-integrityMimi Zohar2020-01-237-1/+220
| |\ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | From patch set cover letter: The IMA subsystem supports measuring asymmetric keys when the key is created or updated[1]. But keys created or updated before a custom IMA policy is loaded are currently not measured. This includes keys added, for instance, to either the .ima or .builtin_trusted_keys keyrings, which happens early in the boot process. Measuring the early boot keys, by design, requires loading a custom IMA policy. This change adds support for queuing keys created or updated before a custom IMA policy is loaded. The queued keys are processed when a custom policy is loaded. Keys created or updated after a custom policy is loaded are measured immediately (not queued). In the case when a custom policy is not loaded within 5 minutes of IMA initialization, the queued keys are freed. [1] https://lore.kernel.org/linux-integrity/20191211164707.4698-1-nramas@linux.microsoft.com/
| | * | IMA: Defined delayed workqueue to free the queued keysLakshmi Ramasubramanian2020-01-233-6/+48
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Keys queued for measurement should be freed if a custom IMA policy was not loaded. Otherwise, the keys will remain queued forever consuming kernel memory. This patch defines a delayed workqueue to handle the above scenario. The workqueue handler is setup to execute 5 minutes after IMA initialization is completed. If a custom IMA policy is loaded before the workqueue handler is scheduled to execute, the workqueue task is cancelled and any queued keys are processed for measurement. But if a custom policy was not loaded then the queued keys are just freed when the delayed workqueue handler is run. Signed-off-by: Lakshmi Ramasubramanian <nramas@linux.microsoft.com> Reported-by: kernel test robot <rong.a.chen@intel.com> # sleeping function called from invalid context Reported-by: kbuild test robot <lkp@intel.com> # redefinition of ima_init_key_queue() function. Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
| | * | IMA: Call workqueue functions to measure queued keysLakshmi Ramasubramanian2020-01-232-0/+11
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Measuring keys requires a custom IMA policy to be loaded. Keys should be queued for measurement if a custom IMA policy is not yet loaded. Keys queued for measurement, if any, should be processed when a custom policy is loaded. This patch updates the IMA hook function ima_post_key_create_or_update() to queue the key if a custom IMA policy has not yet been loaded. And, ima_update_policy() function, which is called when a custom IMA policy is loaded, is updated to process queued keys. Signed-off-by: Lakshmi Ramasubramanian <nramas@linux.microsoft.com> Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
| | * | IMA: Define workqueue for early boot key measurementsLakshmi Ramasubramanian2020-01-234-0/+166
| |/ / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Measuring keys requires a custom IMA policy to be loaded. Keys created or updated before a custom IMA policy is loaded should be queued and will be processed after a custom policy is loaded. This patch defines a workqueue for queuing keys when a custom IMA policy has not yet been loaded. An intermediate Kconfig boolean option namely IMA_QUEUE_EARLY_BOOT_KEYS is used to declare the workqueue functions. A flag namely ima_process_keys is used to check if the key should be queued or should be processed immediately. Signed-off-by: Lakshmi Ramasubramanian <nramas@linux.microsoft.com> Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
| * | IMA: pre-allocate buffer to hold keyrings stringLakshmi Ramasubramanian2020-01-221-8/+30
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | ima_match_keyring() is called while holding rcu read lock. Since this function executes in atomic context, it should not call any function that can sleep (such as kstrdup()). This patch pre-allocates a buffer to hold the keyrings string read from the IMA policy and uses that to match the given keyring. Signed-off-by: Lakshmi Ramasubramanian <nramas@linux.microsoft.com> Fixes: e9085e0ad38a ("IMA: Add support to limit measuring keys") Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
| * | ima: ima/lsm policy rule loading logic bug fixesJanne Karhunen2020-01-221-18/+26
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Keep the ima policy rules around from the beginning even if they appear invalid at the time of loading, as they may become active after an lsm policy load. However, loading a custom IMA policy with unknown LSM labels is only safe after we have transitioned from the "built-in" policy rules to a custom IMA policy. Patch also fixes the rule re-use during the lsm policy reload and makes some prints a bit more human readable. Changelog: v4: - Do not allow the initial policy load refer to non-existing lsm rules. v3: - Fix too wide policy rule matching for non-initialized LSMs v2: - Fix log prints Fixes: b16942455193 ("ima: use the lsm policy update notifier") Cc: Casey Schaufler <casey@schaufler-ca.com> Reported-by: Mimi Zohar <zohar@linux.ibm.com> Signed-off-by: Janne Karhunen <janne.karhunen@gmail.com> Signed-off-by: Konsta Karsisto <konsta.karsisto@gmail.com> Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
| * | ima: add the ability to query the cached hash of a given fileFlorent Revest2020-01-222-0/+55
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This allows other parts of the kernel (perhaps a stacked LSM allowing system monitoring, eg. the proposed KRSI LSM [1]) to retrieve the hash of a given file from IMA if it's present in the iint cache. It's true that the existence of the hash means that it's also in the audit logs or in /sys/kernel/security/ima/ascii_runtime_measurements, but it can be difficult to pull that information out for every subsequent exec. This is especially true if a given host has been up for a long time and the file was first measured a long time ago. It should be kept in mind that this function gives access to cached entries which can be removed, for instance on security_inode_free(). This is based on Peter Moody's patch: https://sourceforge.net/p/linux-ima/mailman/message/33036180/ [1] https://lkml.org/lkml/2019/9/10/393 Signed-off-by: Florent Revest <revest@google.com> Reviewed-by: KP Singh <kpsingh@chromium.org> Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
| * | ima: Add a space after printing LSM rules for readabilityClay Chang2020-01-221-0/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | When reading ima_policy from securityfs, there is a missing space between output string of LSM rules and the remaining rules. Signed-off-by: Clay Chang <clayc@hpe.com> Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
| * | IMA: fix measuring asymmetric keys KconfigLakshmi Ramasubramanian2020-01-093-3/+9
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | As a result of the asymmetric public keys subtype Kconfig option being defined as tristate, with the existing IMA Makefile, ima_asymmetric_keys.c could be built as a kernel module. To prevent this from happening, this patch defines and uses an intermediate Kconfig boolean option named IMA_MEASURE_ASYMMETRIC_KEYS. Signed-off-by: Lakshmi Ramasubramanian <nramas@linux.microsoft.com> Suggested-by: James.Bottomley <James.Bottomley@HansenPartnership.com> Cc: David Howells <dhowells@redhat.com> Cc: Jarkko Sakkinen <jarkko.sakkinen@linux.intel.com> Reported-by: kbuild test robot <lkp@intel.com> # ima_asymmetric_keys.c is built as a kernel module. Fixes: 88e70da170e8 ("IMA: Define an IMA hook to measure keys") Fixes: cb1aa3823c92 ("KEYS: Call the IMA hook to measure keys") [zohar@linux.ibm.com: updated patch description] Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
| * | Merge branch 'next-integrity.measure-keys' into next-integrityMimi Zohar2019-12-1210-20/+207
| |\ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | From "KEYS: Measure keys when they are created or updated" cover letter: Keys created or updated in the system are currently not measured. Therefore an attestation service, for instance, would not be able to attest whether or not the trusted keys keyring(s), for instance, contain only known good (trusted) keys. IMA measures system files, command line arguments passed to kexec, boot aggregate, etc. It can be used to measure keys as well. But there is no mechanism available in the kernel for IMA to know when a key is created or updated. This change aims to address measuring keys created or updated in the system. To achieve the above the following changes have been made: - Added a new IMA hook namely, ima_post_key_create_or_update, which measures the key. This IMA hook is called from key_create_or_update function. The key measurement can be controlled through IMA policy. A new IMA policy function KEY_CHECK has been added to measure keys. "keyrings=" option can be specified for KEY_CHECK to limit measuring the keys loaded onto the specified keyrings only. uid can be specified to further restrict key measurement for keys created by specific user. # measure keys loaded onto any keyring measure func=KEY_CHECK # measure keys loaded onto the IMA keyring only for root user measure func=KEY_CHECK uid=0 keyring=".ima" # measure keys on the BUILTIN and IMA keyrings into a different PCR measure func=KEY_CHECK keyring=".builtin_trusted_keys|.ima" pcr=11
| | * | IMA: Read keyrings= option from the IMA policyLakshmi Ramasubramanian2019-12-121-1/+28
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Read "keyrings=" option, if specified in the IMA policy, and store in the list of IMA rules when the configured IMA policy is read. This patch defines a new policy token enum namely Opt_keyrings and an option flag IMA_KEYRINGS for reading "keyrings=" option from the IMA policy. Updated ima_parse_rule() to parse "keyrings=" option in the policy. Updated ima_policy_show() to display "keyrings=" option. The following example illustrates how key measurement can be verified. Sample "key" measurement rule in the IMA policy: measure func=KEY_CHECK uid=0 keyrings=.ima|.evm template=ima-buf Display "key" measurement in the IMA measurement list: cat /sys/kernel/security/ima/ascii_runtime_measurements 10 faf3...e702 ima-buf sha256:27c915b8ddb9fae7214cf0a8a7043cc3eeeaa7539bcb136f8427067b5f6c3b7b .ima 308202863082...4aee Verify "key" measurement data for a key added to ".ima" keyring: cat /sys/kernel/security/integrity/ima/ascii_runtime_measurements | grep -m 1 "\.ima" | cut -d' ' -f 6 | xxd -r -p |tee ima-cert.der | sha256sum | cut -d' ' -f 1 The output of the above command should match the template hash of the first "key" measurement entry in the IMA measurement list for the key added to ".ima" keyring. The file namely "ima-cert.der" generated by the above command should be a valid x509 certificate (in DER format) and should match the one that was used to import the key to the ".ima" keyring. The certificate file can be verified using openssl tool. Signed-off-by: Lakshmi Ramasubramanian <nramas@linux.microsoft.com> Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
| | * | IMA: Add support to limit measuring keysLakshmi Ramasubramanian2019-12-127-18/+91
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Limit measuring keys to those keys being loaded onto a given set of keyrings only and when the user id (uid) matches if uid is specified in the policy. This patch defines a new IMA policy option namely "keyrings=" that can be used to specify a set of keyrings. If this option is specified in the policy for "measure func=KEY_CHECK" then only the keys loaded onto a keyring given in the "keyrings=" option are measured. If uid is specified in the policy then the key is measured only if the current user id matches the one specified in the policy. Added a new parameter namely "keyring" (name of the keyring) to process_buffer_measurement(). The keyring name is passed to ima_get_action() to determine the required action. ima_match_rules() is updated to check keyring in the policy, if specified, for KEY_CHECK function. Signed-off-by: Lakshmi Ramasubramanian <nramas@linux.microsoft.com> Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
| | * | KEYS: Call the IMA hook to measure keysLakshmi Ramasubramanian2019-12-122-0/+24
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Call the IMA hook from key_create_or_update() function to measure the payload when a new key is created or an existing key is updated. This patch adds the call to the IMA hook from key_create_or_update() function to measure the key on key create or update. Signed-off-by: Lakshmi Ramasubramanian <nramas@linux.microsoft.com> Cc: David Howells <dhowells@redhat.com> Cc: Jarkko Sakkinen <jarkko.sakkinen@linux.intel.com> Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
| | * | IMA: Define an IMA hook to measure keysLakshmi Ramasubramanian2019-12-122-0/+53
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Measure asymmetric keys used for verifying file signatures, certificates, etc. This patch defines a new IMA hook namely ima_post_key_create_or_update() to measure the payload used to create a new asymmetric key or update an existing asymmetric key. Asymmetric key structure is defined only when CONFIG_ASYMMETRIC_PUBLIC_KEY_SUBTYPE is defined. Since the IMA hook measures asymmetric keys, the IMA hook is defined in a new file namely ima_asymmetric_keys.c which is built only if CONFIG_ASYMMETRIC_PUBLIC_KEY_SUBTYPE is defined. Signed-off-by: Lakshmi Ramasubramanian <nramas@linux.microsoft.com> Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
| | * | IMA: Add KEY_CHECK func to measure keysLakshmi Ramasubramanian2019-12-123-2/+9
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Measure keys loaded onto any keyring. This patch defines a new IMA policy func namely KEY_CHECK to measure keys. Updated ima_match_rules() to check for KEY_CHECK and ima_parse_rule() to handle KEY_CHECK. Signed-off-by: Lakshmi Ramasubramanian <nramas@linux.microsoft.com> Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
| | * | IMA: Check IMA policy flagLakshmi Ramasubramanian2019-12-121-0/+3
| |/ / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | process_buffer_measurement() may be called prior to IMA being initialized (for instance, when the IMA hook is called when a key is added to the .builtin_trusted_keys keyring), which would result in a kernel panic. This patch adds the check in process_buffer_measurement() to return immediately if IMA is not initialized yet. Signed-off-by: Lakshmi Ramasubramanian <nramas@linux.microsoft.com> Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
| * | ima: avoid appraise error for hash calc interruptPatrick Callaghan2019-12-121-1/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The integrity_kernel_read() call in ima_calc_file_hash_tfm() can return a value of 0 before all bytes of the file are read. A value of 0 would normally indicate an EOF. This has been observed if a user process is causing a file appraisal and is terminated with a SIGTERM signal. The most common occurrence of seeing the problem is if a shutdown or systemd reload is initiated while files are being appraised. The problem is similar to commit <f5e1040196db> (ima: always return negative code for error) that fixed the problem in ima_calc_file_hash_atfm(). Suggested-by: Mimi Zohar <zohar@linux.ibm.com> Signed-off-by: Patrick Callaghan <patrickc@linux.ibm.com> Reviewed-by: Sascha Hauer <s.hauer@pengutronix.de> Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
* | | Merge tag 'tomoyo-pr-20200128' of git://git.osdn.net/gitroot/tomoyo/tomoyo-test1Linus Torvalds2020-01-291-7/+4
|\ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | Pull tomoyo update from Tetsuo Handa: "One 'int' -> 'atomic_t' conversion patch to suppress KCSAN's warning" * tag 'tomoyo-pr-20200128' of git://git.osdn.net/gitroot/tomoyo/tomoyo-test1: tomoyo: Use atomic_t for statistics counter
| * | | tomoyo: Use atomic_t for statistics counterTetsuo Handa2020-01-021-7/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | syzbot is reporting that there is a race at tomoyo_stat_update() [1]. Although it is acceptable to fail to track exact number of times policy was updated, convert to atomic_t because this is not a hot path. [1] https://syzkaller.appspot.com/bug?id=a4d7b973972eeed410596e6604580e0133b0fc04 Reported-by: syzbot <syzbot+efea72d4a0a1d03596cd@syzkaller.appspotmail.com> Signed-off-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
* | | | Merge tag 's390-5.6-1' of ↵Linus Torvalds2020-01-2913-93/+172
|\ \ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux Pull s390 updates from Vasily Gorbik: - Add clang 10 build support. - Fix BUG() implementation to contain precise bug address, which is relevant for kprobes. - Make ftraced function appear in a stacktrace. - Minor perf improvements and refactoring. - Possible deadlock and recovery fixes in pci code. * tag 's390-5.6-1' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux: s390: fix __EMIT_BUG() macro s390/ftrace: generate traced function stack frame s390: adjust -mpacked-stack support check for clang 10 s390/jump_label: use "i" constraint for clang s390/cpum_sf: Use DIV_ROUND_UP s390/cpum_sf: Use kzalloc and minor changes s390/cpum_sf: Convert debug trace to common layout s390/pci: Fix possible deadlock in recover_store() s390/pci: Recover handle in clp_set_pci_fn()
| * | | | s390: fix __EMIT_BUG() macroSven Schnelle2020-01-225-16/+46
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Setting a kprobe on getname_flags() failed: $ echo 'p:tmr1 getname_flags +0(%r2):ustring' > kprobe_events -bash: echo: write error: Invalid argument Debugging the kprobes code showed that the address of getname_flags() is contained in the __bug_table. Kprobes doesn't allow to set probes at BUG() locations. $ objdump -j __bug_table -x build/fs/namei.o [..] 0000000000000108 R_390_PC32 .text+0x00000000000075a8 000000000000010c R_390_PC32 .L223+0x0000000000000004 I was expecting getname_flags() to start with a BUG(), but: 7598: e3 20 10 00 00 04 lg %r2,0(%r1) 759e: c0 f4 00 00 00 00 jg 759e <putname+0x7e> 75a0: R_390_PLT32DBL kmem_cache_free+0x2 75a4: a7 f4 00 01 j 75a6 <putname+0x86> 00000000000075a8 <getname_flags>: 75a8: c0 04 00 00 00 00 brcl 0,75a8 <getname_flags> 75ae: eb 6f f0 48 00 24 stmg %r6,%r15,72(%r15) 75b4: b9 04 00 ef lgr %r14,%r15 75b8: e3 f0 ff a8 ff 71 lay %r15,-88(%r15) So the BUG() is actually the last opcode of the previous function. Fix this by switching to using the MONITOR CALL (MC) instruction, and set the entry in __bug_table to the beginning of that MC. Reviewed-by: Heiko Carstens <heiko.carstens@de.ibm.com> Signed-off-by: Sven Schnelle <svens@linux.ibm.com> Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
| * | | | s390/ftrace: generate traced function stack frameVasily Gorbik2020-01-221-1/+14
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Currently backtrace from ftraced function does not contain ftraced function itself. e.g. for "path_openat": arch_stack_walk+0x15c/0x2d8 stack_trace_save+0x50/0x68 stack_trace_call+0x15e/0x3d8 ftrace_graph_caller+0x0/0x1c <-- ftrace code do_filp_open+0x7c/0xe8 <-- ftraced function caller do_open_execat+0x76/0x1b8 open_exec+0x52/0x78 load_elf_binary+0x180/0x1160 search_binary_handler+0x8e/0x288 load_script+0x2a8/0x2b8 search_binary_handler+0x8e/0x288 __do_execve_file.isra.39+0x6fa/0xb40 __s390x_sys_execve+0x56/0x68 system_call+0xdc/0x2d8 Ftraced function is expected in the backtrace by ftrace kselftests, which are now failing. It would also be nice to have it for clarity reasons. "ftrace_caller" itself is called without stack frame allocated for it and does not store its caller (ftraced function). Instead it simply allocates a stack frame for "ftrace_trace_function" and sets backchain to point to ftraced function stack frame (which contains ftraced function caller in saved r14). To fix this issue make "ftrace_caller" allocate a stack frame for itself just to store ftraced function for the stack unwinder. As a result backtrace looks like the following: arch_stack_walk+0x15c/0x2d8 stack_trace_save+0x50/0x68 stack_trace_call+0x15e/0x3d8 ftrace_graph_caller+0x0/0x1c <-- ftrace code path_openat+0x6/0xd60 <-- ftraced function do_filp_open+0x7c/0xe8 <-- ftraced function caller do_open_execat+0x76/0x1b8 open_exec+0x52/0x78 load_elf_binary+0x180/0x1160 search_binary_handler+0x8e/0x288 load_script+0x2a8/0x2b8 search_binary_handler+0x8e/0x288 __do_execve_file.isra.39+0x6fa/0xb40 __s390x_sys_execve+0x56/0x68 system_call+0xdc/0x2d8 Reported-by: Sven Schnelle <sven.schnelle@ibm.com> Tested-by: Sven Schnelle <sven.schnelle@ibm.com> Reviewed-by: Heiko Carstens <heiko.carstens@de.ibm.com> Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
| * | | | s390: adjust -mpacked-stack support check for clang 10Vasily Gorbik2020-01-221-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | clang 10 introduces -mpacked-stack compiler option implementation. At the same time currently it does not support a combination of -mpacked-stack and -mbackchain. This leads to the following build error: clang: error: unsupported option '-mpacked-stack with -mbackchain' for target 's390x-ibm-linux' If/when clang adds support for a combination of -mpacked-stack and -mbackchain it would also require -msoft-float (like gcc does). According to Ulrich Weigand "stack slot assigned to the kernel backchain overlaps the stack slot assigned to the FPR varargs (both are required to be placed immediately after the saved r15 slot if present)." Extend -mpacked-stack compiler option support check to include all 3 options -mpacked-stack -mbackchain -msoft-float which must present to support -mpacked-stack with -mbackchain. Acked-by: Heiko Carstens <heiko.carstens@de.ibm.com> Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
| * | | | s390/jump_label: use "i" constraint for clangVasily Gorbik2020-01-221-1/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Currently kernel build fails under clang if jump labels are enabled. The problem is "X" constraint usage "Any operand whatsoever is allowed", for which clang produces the following: .pushsection __jump_table,"aw" .balign 8 .long 0b-.,.Ltmp577-. .quad %r0+0-. # %r0 is not allowed here .popsection Under gcc constraints "X" or "jdd" (gcc > 9) are used for static keys. Ideally, we'd have used "i" for gcc, but it doesn't work in all cases with -fPIC code. This is gcc-specific problem that doesn't exist in llvm. Since clang does not have "jdd" simply always use "i" constraint for it. Suggested-by: Ulrich Weigand <ulrich.weigand@de.ibm.com> Acked-by: Heiko Carstens <heiko.carstens@de.ibm.com> Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
| * | | | s390/cpum_sf: Use DIV_ROUND_UPThomas Richter2020-01-221-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Use macro DIV_ROUND_UP() for calculation of number of SDBT SDBT pages required for index pages. This macro is already used throughout the file. Signed-off-by: Thomas Richter <tmricht@linux.ibm.com> Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
| * | | | s390/cpum_sf: Use kzalloc and minor changesThomas Richter2020-01-221-5/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Use kzalloc() to allocate auxiliary buffer structure initialized with all zeroes to avoid random value in trace output. Avoid double access to SBD hardware flags. Signed-off-by: Thomas Richter <tmricht@linux.ibm.com> Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
| * | | | s390/cpum_sf: Convert debug trace to common layoutThomas Richter2020-01-221-26/+30
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Convert debug traces to print the head/alert/empty marks consistently as decimal numbers. Add some trace statements to enable easier debugging during auxiliary tracing. Signed-off-by: Thomas Richter <tmricht@linux.ibm.com> Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
generated by cgit v1.2.3 (git 2.39.1) at 2025-01-03 07:53:59 +0100