user/sven/linux.git/include/linux, branch v4.14.16

bpf: avoid false sharing of map refcount with max_entries

2018-01-31T13:03:50Z

[ upstream commit be95a845cc4402272994ce290e3ad928aff06cb9 ] In addition to commit b2157399cc98 ("bpf: prevent out-of-bounds speculation") also change the layout of struct bpf_map such that false sharing of fast-path members like max_entries is avoided when the maps reference counter is altered. Therefore enforce them to be placed into separate cachelines. pahole dump after change: struct bpf_map { const struct bpf_map_ops * ops; /* 0 8 */ struct bpf_map * inner_map_meta; /* 8 8 */ void * security; /* 16 8 */ enum bpf_map_type map_type; /* 24 4 */ u32 key_size; /* 28 4 */ u32 value_size; /* 32 4 */ u32 max_entries; /* 36 4 */ u32 map_flags; /* 40 4 */ u32 pages; /* 44 4 */ u32 id; /* 48 4 */ int numa_node; /* 52 4 */ bool unpriv_array; /* 56 1 */ /* XXX 7 bytes hole, try to pack */ /* --- cacheline 1 boundary (64 bytes) --- */ struct user_struct * user; /* 64 8 */ atomic_t refcnt; /* 72 4 */ atomic_t usercnt; /* 76 4 */ struct work_struct work; /* 80 32 */ char name[16]; /* 112 16 */ /* --- cacheline 2 boundary (128 bytes) --- */ /* size: 128, cachelines: 2, members: 17 */ /* sum members: 121, holes: 1, sum holes: 7 */ }; Now all entries in the first cacheline are read only throughout the life time of the map, set up once during map creation. Overall struct size and number of cachelines doesn't change from the reordering. struct bpf_map is usually first member and embedded in map structs in specific map implementations, so also avoid those members to sit at the end where it could potentially share the cacheline with first map values e.g. in the array since remote CPUs could trigger map updates just as well for those (easily dirtying members like max_entries intentionally as well) while having subsequent values in cache. Quoting from Google's Project Zero blog [1]: Additionally, at least on the Intel machine on which this was tested, bouncing modified cache lines between cores is slow, apparently because the MESI protocol is used for cache coherence [8]. Changing the reference counter of an eBPF array on one physical CPU core causes the cache line containing the reference counter to be bounced over to that CPU core, making reads of the reference counter on all other CPU cores slow until the changed reference counter has been written back to memory. Because the length and the reference counter of an eBPF array are stored in the same cache line, this also means that changing the reference counter on one physical CPU core causes reads of the eBPF array's length to be slow on other physical CPU cores (intentional false sharing). While this doesn't 'control' the out-of-bounds speculation through masking the index as in commit b2157399cc98, triggering a manipulation of the map's reference counter is really trivial, so lets not allow to easily affect max_entries from it. Splitting to separate cachelines also generally makes sense from a performance perspective anyway in that fast-path won't have a cache miss if the map gets pinned, reused in other progs, etc out of control path, thus also avoids unintentional false sharing. [1] https://googleprojectzero.blogspot.ch/2018/01/reading-privileged-memory-with-side.html Signed-off-by: Daniel Borkmann Signed-off-by: Alexei Starovoitov Signed-off-by: Greg Kroah-Hartman

net/mlx5: Fix get vector affinity helper function

2018-01-31T13:03:46Z

[ Upstream commit 05e0cc84e00c54fb152d1f4b86bc211823a83d0c ] mlx5_get_vector_affinity used to call pci_irq_get_affinity and after reverting the patch that sets the device affinity via PCI_IRQ_AFFINITY API, calling pci_irq_get_affinity becomes useless and it breaks RDMA mlx5 users. To fix this, this patch provides an alternative way to retrieve IRQ vector affinity using legacy IRQ API, following smp_affinity read procfs implementation. Fixes: 231243c82793 ("Revert mlx5: move affinity hints assignments to generic code") Fixes: a435393acafb ("mlx5: move affinity hints assignments to generic code") Cc: Sagi Grimberg Signed-off-by: Saeed Mahameed Signed-off-by: Greg Kroah-Hartman

{net,ib}/mlx5: Don't disable local loopback multicast traffic when needed

2018-01-31T13:03:46Z

[ Upstream commit 8978cc921fc7fad3f4d6f91f1da01352aeeeff25 ] There are systems platform information management interfaces (such as HOST2BMC) for which we cannot disable local loopback multicast traffic. Separate disable_local_lb_mc and disable_local_lb_uc capability bits so driver will not disable multicast loopback traffic if not supported. (It is expected that Firmware will not set disable_local_lb_mc if HOST2BMC is running for example.) Function mlx5_nic_vport_update_local_lb will do best effort to disable/enable UC/MC loopback traffic and return success only in case it succeeded to changed all allowed by Firmware. Adapt mlx5_ib and mlx5e to support the new cap bits. Fixes: 2c43c5a036be ("net/mlx5e: Enable local loopback in loopback selftest") Fixes: c85023e153e3 ("IB/mlx5: Add raw ethernet local loopback support") Fixes: bded747bb432 ("net/mlx5: Add raw ethernet local loopback firmware command") Signed-off-by: Eran Ben Elisha Cc: kernel-team@fb.com Signed-off-by: Saeed Mahameed Signed-off-by: Greg Kroah-Hartman

Revert "module: Add retpoline tag to VERMAGIC"

2018-01-31T13:03:41Z

commit 5132ede0fe8092b043dae09a7cc32b8ae7272baa upstream. This reverts commit 6cfb521ac0d5b97470883ff9b7facae264b7ab12. Turns out distros do not want to make retpoline as part of their "ABI", so this patch should not have been merged. Sorry Andi, this was my fault, I suggested it when your original patch was the "correct" way of doing this instead. Reported-by: Jiri Kosina Fixes: 6cfb521ac0d5 ("module: Add retpoline tag to VERMAGIC") Acked-by: Andi Kleen Cc: Thomas Gleixner Cc: David Woodhouse Cc: rusty@rustcorp.com.au Cc: arjan.van.de.ven@intel.com Cc: jeyu@kernel.org Signed-off-by: Linus Torvalds Signed-off-by: Greg Kroah-Hartman

mm, page_vma_mapped: Drop faulty pointer arithmetics in check_pte()

2018-01-23T18:58:21Z

commit 0d665e7b109d512b7cae3ccef6e8654714887844 upstream. Tetsuo reported random crashes under memory pressure on 32-bit x86 system and tracked down to change that introduced page_vma_mapped_walk(). The root cause of the issue is the faulty pointer math in check_pte(). As ->pte may point to an arbitrary page we have to check that they are belong to the section before doing math. Otherwise it may lead to weird results. It wasn't noticed until now as mem_map[] is virtually contiguous on flatmem or vmemmap sparsemem. Pointer arithmetic just works against all 'struct page' pointers. But with classic sparsemem, it doesn't because each section memap is allocated separately and so consecutive pfns crossing two sections might have struct pages at completely unrelated addresses. Let's restructure code a bit and replace pointer arithmetic with operations on pfns. Signed-off-by: Kirill A. Shutemov Reported-and-tested-by: Tetsuo Handa Acked-by: Michal Hocko Fixes: ace71a19cec5 ("mm: introduce page_vma_mapped_walk()") Signed-off-by: Linus Torvalds Signed-off-by: Greg Kroah-Hartman

module: Add retpoline tag to VERMAGIC

2018-01-23T18:58:14Z

commit 6cfb521ac0d5b97470883ff9b7facae264b7ab12 upstream. Add a marker for retpoline to the module VERMAGIC. This catches the case when a non RETPOLINE compiled module gets loaded into a retpoline kernel, making it insecure. It doesn't handle the case when retpoline has been runtime disabled. Even in this case the match of the retcompile status will be enforced. This implies that even with retpoline run time disabled all modules loaded need to be recompiled. Signed-off-by: Andi Kleen Signed-off-by: Thomas Gleixner Reviewed-by: Greg Kroah-Hartman Acked-by: David Woodhouse Cc: rusty@rustcorp.com.au Cc: arjan.van.de.ven@intel.com Cc: jeyu@kernel.org Cc: torvalds@linux-foundation.org Link: https://lkml.kernel.org/r/20180116205228.4890-1-andi@firstfloor.org Signed-off-by: Greg Kroah-Hartman

delayacct: Account blkio completion on the correct task

2018-01-23T18:58:13Z

commit c96f5471ce7d2aefd0dda560cc23f08ab00bc65d upstream. Before commit: e33a9bba85a8 ("sched/core: move IO scheduling accounting from io_schedule_timeout() into scheduler") delayacct_blkio_end() was called after context-switching into the task which completed I/O. This resulted in double counting: the task would account a delay both waiting for I/O and for time spent in the runqueue. With e33a9bba85a8, delayacct_blkio_end() is called by try_to_wake_up(). In ttwu, we have not yet context-switched. This is more correct, in that the delay accounting ends when the I/O is complete. But delayacct_blkio_end() relies on 'get_current()', and we have not yet context-switched into the task whose I/O completed. This results in the wrong task having its delay accounting statistics updated. Instead of doing that, pass the task_struct being woken to delayacct_blkio_end(), so that it can update the statistics of the correct task. Signed-off-by: Josh Snyder Acked-by: Tejun Heo Acked-by: Balbir Singh Cc: Brendan Gregg Cc: Jens Axboe Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: linux-block@vger.kernel.org Fixes: e33a9bba85a8 ("sched/core: move IO scheduling accounting from io_schedule_timeout() into scheduler") Link: http://lkml.kernel.org/r/1513613712-571-1-git-send-email-joshs@netflix.com Signed-off-by: Ingo Molnar Signed-off-by: Greg Kroah-Hartman

sysfs/cpu: Add vulnerability folder

2018-01-17T08:45:27Z

commit 87590ce6e373d1a5401f6539f0c59ef92dd924a9 upstream. As the meltdown/spectre problem affects several CPU architectures, it makes sense to have common way to express whether a system is affected by a particular vulnerability or not. If affected the way to express the mitigation should be common as well. Create /sys/devices/system/cpu/vulnerabilities folder and files for meltdown, spectre_v1 and spectre_v2. Allow architectures to override the show function. Signed-off-by: Thomas Gleixner Reviewed-by: Greg Kroah-Hartman Reviewed-by: Konrad Rzeszutek Wilk Cc: Peter Zijlstra Cc: Will Deacon Cc: Dave Hansen Cc: Linus Torvalds Cc: Borislav Petkov Cc: David Woodhouse Link: https://lkml.kernel.org/r/20180107214913.096657732@linutronix.de Signed-off-by: Greg Kroah-Hartman

kdump: write correct address of mem_section into vmcoreinfo

2018-01-17T08:45:27Z

commit a0b1280368d1e91ab72f849ef095b4f07a39bbf1 upstream. Depending on configuration mem_section can now be an array or a pointer to an array allocated dynamically. In most cases, we can continue to refer to it as 'mem_section' regardless of what it is. But there's one exception: '&mem_section' means "address of the array" if mem_section is an array, but if mem_section is a pointer, it would mean "address of the pointer". We've stepped onto this in kdump code. VMCOREINFO_SYMBOL(mem_section) writes down address of pointer into vmcoreinfo, not array as we wanted. Let's introduce VMCOREINFO_SYMBOL_ARRAY() that would handle the situation correctly for both cases. Link: http://lkml.kernel.org/r/20180112162532.35896-1-kirill.shutemov@linux.intel.com Signed-off-by: Kirill A. Shutemov Fixes: 83e3c48729d9 ("mm/sparsemem: Allocate mem_section at runtime for CONFIG_SPARSEMEM_EXTREME=y") Acked-by: Baoquan He Acked-by: Dave Young Cc: Ingo Molnar Cc: Thomas Gleixner Cc: "H. Peter Anvin" Cc: Greg Kroah-Hartman Cc: Dave Young Cc: Baoquan He Cc: Vivek Goyal Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds Signed-off-by: Greg Kroah-Hartman

bpf: prevent out-of-bounds speculation

2018-01-17T08:45:25Z

commit b2157399cc9898260d6031c5bfe45fe137c1fbe7 upstream. Under speculation, CPUs may mis-predict branches in bounds checks. Thus, memory accesses under a bounds check may be speculated even if the bounds check fails, providing a primitive for building a side channel. To avoid leaking kernel data round up array-based maps and mask the index after bounds check, so speculated load with out of bounds index will load either valid value from the array or zero from the padded area. Unconditionally mask index for all array types even when max_entries are not rounded to power of 2 for root user. When map is created by unpriv user generate a sequence of bpf insns that includes AND operation to make sure that JITed code includes the same 'index & index_mask' operation. If prog_array map is created by unpriv user replace bpf_tail_call(ctx, map, index); with if (index >= max_entries) { index &= map->index_mask; bpf_tail_call(ctx, map, index); } (along with roundup to power 2) to prevent out-of-bounds speculation. There is secondary redundant 'if (index >= max_entries)' in the interpreter and in all JITs, but they can be optimized later if necessary. Other array-like maps (cpumap, devmap, sockmap, perf_event_array, cgroup_array) cannot be used by unpriv, so no changes there. That fixes bpf side of "Variant 1: bounds check bypass (CVE-2017-5753)" on all architectures with and without JIT. v2->v3: Daniel noticed that attack potentially can be crafted via syscall commands without loading the program, so add masking to those paths as well. Signed-off-by: Alexei Starovoitov Acked-by: John Fastabend Signed-off-by: Daniel Borkmann Cc: Jiri Slaby Signed-off-by: Greg Kroah-Hartman