| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
Patch series "Multi-Gen LRU Framework", v14.
What's new
==========
1. OpenWrt, in addition to Android, Arch Linux Zen, Armbian, ChromeOS,
Liquorix, post-factum and XanMod, is now shipping MGLRU on 5.15.
2. Fixed long-tailed direct reclaim latency seen on high-memory (TBs)
machines. The old direct reclaim backoff, which tries to enforce a
minimum fairness among all eligible memcgs, over-swapped by about
(total_mem>>DEF_PRIORITY)-nr_to_reclaim. The new backoff, which
pulls the plug on swapping once the target is met, trades some
fairness for curtailed latency:
https://lore.kernel.org/r/20220918080010.2920238-10-yuzhao@google.com/
3. Fixed minior build warnings and conflicts. More comments and nits.
TLDR
====
The current page reclaim is too expensive in terms of CPU usage and it
often makes poor choices about what to evict. This patchset offers an
alternative solution that is performant, versatile and
straightforward.
Patchset overview
=================
The design and implementation overview is in patch 14:
https://lore.kernel.org/r/20220918080010.2920238-15-yuzhao@google.com/
01. mm: x86, arm64: add arch_has_hw_pte_young()
02. mm: x86: add CONFIG_ARCH_HAS_NONLEAF_PMD_YOUNG
Take advantage of hardware features when trying to clear the accessed
bit in many PTEs.
03. mm/vmscan.c: refactor shrink_node()
04. Revert "include/linux/mm_inline.h: fold __update_lru_size() into
its sole caller"
Minor refactors to improve readability for the following patches.
05. mm: multi-gen LRU: groundwork
Adds the basic data structure and the functions that insert pages to
and remove pages from the multi-gen LRU (MGLRU) lists.
06. mm: multi-gen LRU: minimal implementation
A minimal implementation without optimizations.
07. mm: multi-gen LRU: exploit locality in rmap
Exploits spatial locality to improve efficiency when using the rmap.
08. mm: multi-gen LRU: support page table walks
Further exploits spatial locality by optionally scanning page tables.
09. mm: multi-gen LRU: optimize multiple memcgs
Optimizes the overall performance for multiple memcgs running mixed
types of workloads.
10. mm: multi-gen LRU: kill switch
Adds a kill switch to enable or disable MGLRU at runtime.
11. mm: multi-gen LRU: thrashing prevention
12. mm: multi-gen LRU: debugfs interface
Provide userspace with features like thrashing prevention, working set
estimation and proactive reclaim.
13. mm: multi-gen LRU: admin guide
14. mm: multi-gen LRU: design doc
Add an admin guide and a design doc.
Benchmark results
=================
Independent lab results
-----------------------
Based on the popularity of searches [01] and the memory usage in
Google's public cloud, the most popular open-source memory-hungry
applications, in alphabetical order, are:
Apache Cassandra Memcached
Apache Hadoop MongoDB
Apache Spark PostgreSQL
MariaDB (MySQL) Redis
An independent lab evaluated MGLRU with the most widely used benchmark
suites for the above applications. They posted 960 data points along
with kernel metrics and perf profiles collected over more than 500
hours of total benchmark time. Their final reports show that, with 95%
confidence intervals (CIs), the above applications all performed
significantly better for at least part of their benchmark matrices.
On 5.14:
1. Apache Spark [02] took 95% CIs [9.28, 11.19]% and [12.20, 14.93]%
less wall time to sort three billion random integers, respectively,
under the medium- and the high-concurrency conditions, when
overcommitting memory. There were no statistically significant
changes in wall time for the rest of the benchmark matrix.
2. MariaDB [03] achieved 95% CIs [5.24, 10.71]% and [20.22, 25.97]%
more transactions per minute (TPM), respectively, under the medium-
and the high-concurrency conditions, when overcommitting memory.
There were no statistically significant changes in TPM for the rest
of the benchmark matrix.
3. Memcached [04] achieved 95% CIs [23.54, 32.25]%, [20.76, 41.61]%
and [21.59, 30.02]% more operations per second (OPS), respectively,
for sequential access, random access and Gaussian (distribution)
access, when THP=always; 95% CIs [13.85, 15.97]% and
[23.94, 29.92]% more OPS, respectively, for random access and
Gaussian access, when THP=never. There were no statistically
significant changes in OPS for the rest of the benchmark matrix.
4. MongoDB [05] achieved 95% CIs [2.23, 3.44]%, [6.97, 9.73]% and
[2.16, 3.55]% more operations per second (OPS), respectively, for
exponential (distribution) access, random access and Zipfian
(distribution) access, when underutilizing memory; 95% CIs
[8.83, 10.03]%, [21.12, 23.14]% and [5.53, 6.46]% more OPS,
respectively, for exponential access, random access and Zipfian
access, when overcommitting memory.
On 5.15:
5. Apache Cassandra [06] achieved 95% CIs [1.06, 4.10]%, [1.94, 5.43]%
and [4.11, 7.50]% more operations per second (OPS), respectively,
for exponential (distribution) access, random access and Zipfian
(distribution) access, when swap was off; 95% CIs [0.50, 2.60]%,
[6.51, 8.77]% and [3.29, 6.75]% more OPS, respectively, for
exponential access, random access and Zipfian access, when swap was
on.
6. Apache Hadoop [07] took 95% CIs [5.31, 9.69]% and [2.02, 7.86]%
less average wall time to finish twelve parallel TeraSort jobs,
respectively, under the medium- and the high-concurrency
conditions, when swap was on. There were no statistically
significant changes in average wall time for the rest of the
benchmark matrix.
7. PostgreSQL [08] achieved 95% CI [1.75, 6.42]% more transactions per
minute (TPM) under the high-concurrency condition, when swap was
off; 95% CIs [12.82, 18.69]% and [22.70, 46.86]% more TPM,
respectively, under the medium- and the high-concurrency
conditions, when swap was on. There were no statistically
significant changes in TPM for the rest of the benchmark matrix.
8. Redis [09] achieved 95% CIs [0.58, 5.94]%, [6.55, 14.58]% and
[11.47, 19.36]% more total operations per second (OPS),
respectively, for sequential access, random access and Gaussian
(distribution) access, when THP=always; 95% CIs [1.27, 3.54]%,
[10.11, 14.81]% and [8.75, 13.64]% more total OPS, respectively,
for sequential access, random access and Gaussian access, when
THP=never.
Our lab results
---------------
To supplement the above results, we ran the following benchmark suites
on 5.16-rc7 and found no regressions [10].
fs_fio_bench_hdd_mq pft
fs_lmbench pgsql-hammerdb
fs_parallelio redis
fs_postmark stream
hackbench sysbenchthread
kernbench tpcc_spark
memcached unixbench
multichase vm-scalability
mutilate will-it-scale
nginx
[01] https://trends.google.com
[02] https://lore.kernel.org/r/20211102002002.92051-1-bot@edi.works/
[03] https://lore.kernel.org/r/20211009054315.47073-1-bot@edi.works/
[04] https://lore.kernel.org/r/20211021194103.65648-1-bot@edi.works/
[05] https://lore.kernel.org/r/20211109021346.50266-1-bot@edi.works/
[06] https://lore.kernel.org/r/20211202062806.80365-1-bot@edi.works/
[07] https://lore.kernel.org/r/20211209072416.33606-1-bot@edi.works/
[08] https://lore.kernel.org/r/20211218071041.24077-1-bot@edi.works/
[09] https://lore.kernel.org/r/20211122053248.57311-1-bot@edi.works/
[10] https://lore.kernel.org/r/20220104202247.2903702-1-yuzhao@google.com/
Read-world applications
=======================
Third-party testimonials
------------------------
Konstantin reported [11]:
I have Archlinux with 8G RAM + zswap + swap. While developing, I
have lots of apps opened such as multiple LSP-servers for different
langs, chats, two browsers, etc... Usually, my system gets quickly
to a point of SWAP-storms, where I have to kill LSP-servers,
restart browsers to free memory, etc, otherwise the system lags
heavily and is barely usable.
1.5 day ago I migrated from 5.11.15 kernel to 5.12 + the LRU
patchset, and I started up by opening lots of apps to create memory
pressure, and worked for a day like this. Till now I had not a
single SWAP-storm, and mind you I got 3.4G in SWAP. I was never
getting to the point of 3G in SWAP before without a single
SWAP-storm.
Vaibhav from IBM reported [12]:
In a synthetic MongoDB Benchmark, seeing an average of ~19%
throughput improvement on POWER10(Radix MMU + 64K Page Size) with
MGLRU patches on top of 5.16 kernel for MongoDB + YCSB across
three different request distributions, namely, Exponential, Uniform
and Zipfan.
Shuang from U of Rochester reported [13]:
With the MGLRU, fio achieved 95% CIs [38.95, 40.26]%, [4.12, 6.64]%
and [9.26, 10.36]% higher throughput, respectively, for random
access, Zipfian (distribution) access and Gaussian (distribution)
access, when the average number of jobs per CPU is 1; 95% CIs
[42.32, 49.15]%, [9.44, 9.89]% and [20.99, 22.86]% higher
throughput, respectively, for random access, Zipfian access and
Gaussian access, when the average number of jobs per CPU is 2.
Daniel from Michigan Tech reported [14]:
With Memcached allocating ~100GB of byte-addressable Optante,
performance improvement in terms of throughput (measured as queries
per second) was about 10% for a series of workloads.
Large-scale deployments
-----------------------
We've rolled out MGLRU to tens of millions of ChromeOS users and
about a million Android users. Google's fleetwide profiling [15] shows
an overall 40% decrease in kswapd CPU usage, in addition to
improvements in other UX metrics, e.g., an 85% decrease in the number
of low-memory kills at the 75th percentile and an 18% decrease in
app launch time at the 50th percentile.
The downstream kernels that have been using MGLRU include:
1. Android [16]
2. Arch Linux Zen [17]
3. Armbian [18]
4. ChromeOS [19]
5. Liquorix [20]
6. OpenWrt [21]
7. post-factum [22]
8. XanMod [23]
[11] https://lore.kernel.org/r/140226722f2032c86301fbd326d91baefe3d7d23.camel@yandex.ru/
[12] https://lore.kernel.org/r/87czj3mux0.fsf@vajain21.in.ibm.com/
[13] https://lore.kernel.org/r/20220105024423.26409-1-szhai2@cs.rochester.edu/
[14] https://lore.kernel.org/r/CA+4-3vksGvKd18FgRinxhqHetBS1hQekJE2gwco8Ja-bJWKtFw@mail.gmail.com/
[15] https://dl.acm.org/doi/10.1145/2749469.2750392
[16] https://android.com
[17] https://archlinux.org
[18] https://armbian.com
[19] https://chromium.org
[20] https://liquorix.net
[21] https://openwrt.org
[22] https://codeberg.org/pf-kernel
[23] https://xanmod.org
Summary
=======
The facts are:
1. The independent lab results and the real-world applications
indicate substantial improvements; there are no known regressions.
2. Thrashing prevention, working set estimation and proactive reclaim
work out of the box; there are no equivalent solutions.
3. There is a lot of new code; no smaller changes have been
demonstrated similar effects.
Our options, accordingly, are:
1. Given the amount of evidence, the reported improvements will likely
materialize for a wide range of workloads.
2. Gauging the interest from the past discussions, the new features
will likely be put to use for both personal computers and data
centers.
3. Based on Google's track record, the new code will likely be well
maintained in the long term. It'd be more difficult if not
impossible to achieve similar effects with other approaches.
This patch (of 14):
Some architectures automatically set the accessed bit in PTEs, e.g., x86
and arm64 v8.2. On architectures that do not have this capability,
clearing the accessed bit in a PTE usually triggers a page fault following
the TLB miss of this PTE (to emulate the accessed bit).
Being aware of this capability can help make better decisions, e.g.,
whether to spread the work out over a period of time to reduce bursty page
faults when trying to clear the accessed bit in many PTEs.
Note that theoretically this capability can be unreliable, e.g.,
hotplugged CPUs might be different from builtin ones. Therefore it should
not be used in architecture-independent code that involves correctness,
e.g., to determine whether TLB flushes are required (in combination with
the accessed bit).
Link: https://lkml.kernel.org/r/20220918080010.2920238-1-yuzhao@google.com
Link: https://lkml.kernel.org/r/20220918080010.2920238-2-yuzhao@google.com
Signed-off-by: Yu Zhao <yuzhao@google.com>
Reviewed-by: Barry Song <baohua@kernel.org>
Acked-by: Brian Geffon <bgeffon@google.com>
Acked-by: Jan Alexander Steffens (heftig) <heftig@archlinux.org>
Acked-by: Oleksandr Natalenko <oleksandr@natalenko.name>
Acked-by: Steven Barrett <steven@liquorix.net>
Acked-by: Suleiman Souhlal <suleiman@google.com>
Acked-by: Will Deacon <will@kernel.org>
Tested-by: Daniel Byrne <djbyrne@mtu.edu>
Tested-by: Donald Carr <d@chaos-reins.com>
Tested-by: Holger Hoffstätte <holger@applied-asynchrony.com>
Tested-by: Konstantin Kharlamov <Hi-Angel@yandex.ru>
Tested-by: Shuang Zhai <szhai2@cs.rochester.edu>
Tested-by: Sofia Trinh <sofia.trinh@edi.works>
Tested-by: Vaibhav Jain <vaibhav@linux.ibm.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Hillf Danton <hdanton@sina.com>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: linux-arm-kernel@lists.infradead.org
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Michael Larabel <Michael@MichaelLarabel.com>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Mike Rapoport <rppt@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Tejun Heo <tj@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Miaohe Lin <linmiaohe@huawei.com>
Cc: Mike Rapoport <rppt@linux.ibm.com>
Cc: Qi Zheng <zhengqi.arch@bytedance.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
|
|
We used to have swapfile_maximum_size() fetching a maximum value of
swapfile size per-arch.
As the caller of max_swapfile_size() grows, this patch introduce a
variable "swapfile_maximum_size" and cache the value of old
max_swapfile_size(), so that we don't need to calculate the value every
time.
Caching the value in swapfile_init() is safe because when reaching the
phase we should have initialized all the relevant information. Here the
major arch to take care of is x86, which defines the max swapfile size
based on L1TF mitigation.
Here both X86_BUG_L1TF or l1tf_mitigation should have been setup properly
when reaching swapfile_init(). As a reference, the code path looks like
this for x86:
- start_kernel
- setup_arch
- early_cpu_init
- early_identify_cpu --> setup X86_BUG_L1TF
- parse_early_param
- l1tf_cmdline --> set l1tf_mitigation
- check_bugs
- l1tf_select_mitigation --> set l1tf_mitigation
- arch_call_rest_init
- rest_init
- kernel_init
- kernel_init_freeable
- do_basic_setup
- do_initcalls --> calls swapfile_init() (initcall level 4)
The swapfile size only depends on swp pte format on non-x86 archs, so
caching it is safe too.
Since at it, rename max_swapfile_size() to arch_max_swapfile_size()
because arch can define its own function, so it's more straightforward to
have "arch_" as its prefix. At the meantime, export swapfile_maximum_size
to replace the old usages of max_swapfile_size().
[peterx@redhat.com: declare arch_max_swapfile_size) in swapfile.h]
Link: https://lkml.kernel.org/r/YxTh1GuC6ro5fKL5@xz-m1.local
Link: https://lkml.kernel.org/r/20220811161331.37055-7-peterx@redhat.com
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: "Huang, Ying" <ying.huang@intel.com>
Cc: Alistair Popple <apopple@nvidia.com>
Cc: Andi Kleen <andi.kleen@intel.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: David Hildenbrand <david@redhat.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: "Kirill A . Shutemov" <kirill@shutemov.name>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Nadav Amit <nadav.amit@gmail.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Dave Hansen <dave.hansen@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
|
|
Patch series "mm: Remember a/d bits for migration entries", v4.
Problem
=======
When migrating a page, right now we always mark the migrated page as old &
clean.
However that could lead to at least two problems:
(1) We lost the real hot/cold information while we could have persisted.
That information shouldn't change even if the backing page is changed
after the migration,
(2) There can be always extra overhead on the immediate next access to
any migrated page, because hardware MMU needs cycles to set the young
bit again for reads, and dirty bits for write, as long as the
hardware MMU supports these bits.
Many of the recent upstream works showed that (2) is not something trivial
and actually very measurable. In my test case, reading 1G chunk of memory
- jumping in page size intervals - could take 99ms just because of the
extra setting on the young bit on a generic x86_64 system, comparing to
4ms if young set.
This issue is originally reported by Andrea Arcangeli.
Solution
========
To solve this problem, this patchset tries to remember the young/dirty
bits in the migration entries and carry them over when recovering the
ptes.
We have the chance to do so because in many systems the swap offset is not
really fully used. Migration entries use swp offset to store PFN only,
while the PFN is normally not as large as swp offset and normally smaller.
It means we do have some free bits in swp offset that we can use to store
things like A/D bits, and that's how this series tried to approach this
problem.
max_swapfile_size() is used here to detect per-arch offset length in swp
entries. We'll automatically remember the A/D bits when we find that we
have enough swp offset field to keep both the PFN and the extra bits.
Since max_swapfile_size() can be slow, the last two patches cache the
results for it and also swap_migration_ad_supported as a whole.
Known Issues / TODOs
====================
We still haven't taught madvise() to recognize the new A/D bits in
migration entries, namely MADV_COLD/MADV_FREE. E.g. when MADV_COLD upon
a migration entry. It's not clear yet on whether we should clear the A
bit, or we should just drop the entry directly.
We didn't teach idle page tracking on the new migration entries, because
it'll need larger rework on the tree on rmap pgtable walk. However it
should make it already better because before this patchset page will be
old page after migration, so the series will fix potential false negative
of idle page tracking when pages were migrated before observing.
The other thing is migration A/D bits will not start to working for
private device swap entries. The code is there for completeness but since
private device swap entries do not yet have fields to store A/D bits, even
if we'll persistent A/D across present pte switching to migration entry,
we'll lose it again when the migration entry converted to private device
swap entry.
Tests
=====
After the patchset applied, the immediate read access test [1] of above 1G
chunk after migration can shrink from 99ms to 4ms. The test is done by
moving 1G pages from node 0->1->0 then read it in page size jumps. The
test is with Intel(R) Xeon(R) CPU E5-2630 v4 @ 2.20GHz.
Similar effect can also be measured when writting the memory the 1st time
after migration.
After applying the patchset, both initial immediate read/write after page
migrated will perform similarly like before migration happened.
Patch Layout
============
Patch 1-2: Cleanups from either previous versions or on swapops.h macros.
Patch 3-4: Prepare for the introduction of migration A/D bits
Patch 5: The core patch to remember young/dirty bit in swap offsets.
Patch 6-7: Cache relevant fields to make migration_entry_supports_ad() fast.
[1] https://github.com/xzpeter/clibs/blob/master/misc/swap-young.c
This patch (of 7):
Replace all the magic "5" with the macro.
Link: https://lkml.kernel.org/r/20220811161331.37055-1-peterx@redhat.com
Link: https://lkml.kernel.org/r/20220811161331.37055-2-peterx@redhat.com
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Huang Ying <ying.huang@intel.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: "Kirill A . Shutemov" <kirill@shutemov.name>
Cc: Alistair Popple <apopple@nvidia.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Andi Kleen <andi.kleen@intel.com>
Cc: Nadav Amit <nadav.amit@gmail.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Dave Hansen <dave.hansen@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
|
|
Pull x86 fixes from Dave Hansen:
- A performance fix for recent large AMD systems that avoids an ancient
cpu idle hardware workaround
- A new Intel model number. Folks like these upstream as soon as
possible so that each developer doing feature development doesn't
need to carry their own #define
- SGX fixes for a userspace crash and a rare kernel warning
* tag 'x86_urgent_for_v6.0-rc8' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
ACPI: processor idle: Practically limit "Dummy wait" workaround to old Intel systems
x86/sgx: Handle VA page allocation failure for EAUG on PF.
x86/sgx: Do not fail on incomplete sanitization on premature stop of ksgxd
x86/cpu: Add CPU model numbers for Meteor Lake
|
|
Pull last (?) hotfixes from Andrew Morton:
"26 hotfixes.
8 are for issues which were introduced during this -rc cycle, 18 are
for earlier issues, and are cc:stable"
* tag 'mm-hotfixes-stable-2022-09-26' of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm: (26 commits)
x86/uaccess: avoid check_object_size() in copy_from_user_nmi()
mm/page_isolation: fix isolate_single_pageblock() isolation behavior
mm,hwpoison: check mm when killing accessing process
mm/hugetlb: correct demote page offset logic
mm: prevent page_frag_alloc() from corrupting the memory
mm: bring back update_mmu_cache() to finish_fault()
frontswap: don't call ->init if no ops are registered
mm/huge_memory: use pfn_to_online_page() in split_huge_pages_all()
mm: fix madivse_pageout mishandling on non-LRU page
powerpc/64s/radix: don't need to broadcast IPI for radix pmd collapse flush
mm: gup: fix the fast GUP race against THP collapse
mm: fix dereferencing possible ERR_PTR
vmscan: check folio_test_private(), not folio_get_private()
mm: fix VM_BUG_ON in __delete_from_swap_cache()
tools: fix compilation after gfp_types.h split
mm/damon/dbgfs: fix memory leak when using debugfs_lookup()
mm/migrate_device.c: copy pte dirty bit to page
mm/migrate_device.c: add missing flush_cache_page()
mm/migrate_device.c: flush TLB while holding PTL
x86/mm: disable instrumentations of mm/pgprot.c
...
|
|
The check_object_size() helper under CONFIG_HARDENED_USERCOPY is designed
to skip any checks where the length is known at compile time as a
reasonable heuristic to avoid "likely known-good" cases. However, it can
only do this when the copy_*_user() helpers are, themselves, inline too.
Using find_vmap_area() requires taking a spinlock. The
check_object_size() helper can call find_vmap_area() when the destination
is in vmap memory. If show_regs() is called in interrupt context, it will
attempt a call to copy_from_user_nmi(), which may call check_object_size()
and then find_vmap_area(). If something in normal context happens to be
in the middle of calling find_vmap_area() (with the spinlock held), the
interrupt handler will hang forever.
The copy_from_user_nmi() call is actually being called with a fixed-size
length, so check_object_size() should never have been called in the first
place. Given the narrow constraints, just replace the
__copy_from_user_inatomic() call with an open-coded version that calls
only into the sanitizers and not check_object_size(), followed by a call
to raw_copy_from_user().
[akpm@linux-foundation.org: no instrument_copy_from_user() in my tree...]
Link: https://lkml.kernel.org/r/20220919201648.2250764-1-keescook@chromium.org
Link: https://lore.kernel.org/all/CAOUHufaPshtKrTWOz7T7QFYUNVGFm0JBjvM700Nhf9qEL9b3EQ@mail.gmail.com
Fixes: 0aef499f3172 ("mm/usercopy: Detect vmalloc overruns")
Signed-off-by: Kees Cook <keescook@chromium.org>
Reported-by: Yu Zhao <yuzhao@google.com>
Reported-by: Florian Lehner <dev@der-flo.net>
Suggested-by: Andrew Morton <akpm@linux-foundation.org>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Florian Lehner <dev@der-flo.net>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Josh Poimboeuf <jpoimboe@kernel.org>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
|
|
The ZERO_CALL_USED_REGS feature may zero out caller-saved registers
before returning.
In spurious_kernel_fault(), the "pte_offset_kernel()" call results in
this assembly code:
.Ltmp151:
#APP
# ALT: oldnstr
.Ltmp152:
.Ltmp153:
.Ltmp154:
.section .discard.retpoline_safe,"",@progbits
.quad .Ltmp154
.text
callq *pv_ops+536(%rip)
.Ltmp155:
.section .parainstructions,"a",@progbits
.p2align 3, 0x0
.quad .Ltmp153
.byte 67
.byte .Ltmp155-.Ltmp153
.short 1
.text
.Ltmp156:
# ALT: padding
.zero (-(((.Ltmp157-.Ltmp158)-(.Ltmp156-.Ltmp152))>0))*((.Ltmp157-.Ltmp158)-(.Ltmp156-.Ltmp152)),144
.Ltmp159:
.section .altinstructions,"a",@progbits
.Ltmp160:
.long .Ltmp152-.Ltmp160
.Ltmp161:
.long .Ltmp158-.Ltmp161
.short 33040
.byte .Ltmp159-.Ltmp152
.byte .Ltmp157-.Ltmp158
.text
.section .altinstr_replacement,"ax",@progbits
# ALT: replacement 1
.Ltmp158:
movq %rdi, %rax
.Ltmp157:
.text
#NO_APP
.Ltmp162:
testb $-128, %dil
The "testb" here is using %dil, but the %rdi register was cleared before
returning from "callq *pv_ops+536(%rip)". Adding the proper constraints
results in the use of a different register:
movq %r11, %rdi
# Similar to above.
testb $-128, %r11b
Link: https://github.com/KSPP/linux/issues/192
Signed-off-by: Bill Wendling <morbo@google.com>
Reported-and-tested-by: Nathan Chancellor <nathan@kernel.org>
Fixes: 035f7f87b729 ("randstruct: Enable Clang support")
Reviewed-by: Juergen Gross <jgross@suse.com>
Link: https://lore.kernel.org/lkml/fa6df43b-8a1a-8ad1-0236-94d2a0b588fa@suse.com/
Signed-off-by: Kees Cook <keescook@chromium.org>
Link: https://lore.kernel.org/r/20220902213750.1124421-3-morbo@google.com
|
|
Drive-by clean up of the comment.
[ Impact: cleanup]
Signed-off-by: Bill Wendling <morbo@google.com>
Signed-off-by: Kees Cook <keescook@chromium.org>
Link: https://lore.kernel.org/r/20220902213750.1124421-2-morbo@google.com
|
|
Clang produces a false positive when building with CONFIG_FORTIFY_SOURCE=y
and CONFIG_UBSAN_BOUNDS=y when operating on an array with a dynamic
offset. Work around this by using a direct assignment of an empty
instance. Avoids this warning:
../include/linux/fortify-string.h:309:4: warning: call to __write_overflow_field declared with 'warn
ing' attribute: detected write beyond size of field (1st parameter); maybe use struct_group()? [-Wat
tribute-warning]
__write_overflow_field(p_size_field, size);
^
which was isolated to the memset() call in xen_load_idt().
Note that this looks very much like another bug that was worked around:
https://github.com/ClangBuiltLinux/linux/issues/1592
Cc: Juergen Gross <jgross@suse.com>
Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: x86@kernel.org
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: xen-devel@lists.xenproject.org
Reviewed-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Link: https://lore.kernel.org/lkml/41527d69-e8ab-3f86-ff37-6b298c01d5bc@oracle.com
Signed-off-by: Kees Cook <keescook@chromium.org>
|
|
All uses of arch_kprobe_override_function() have been removed by
commit 540adea3809f ("error-injection: Separate error-injection
from kprobe"), so remove the declaration, too.
Link: https://lkml.kernel.org/r/20220914110437.1436353-3-cuigaosheng1@huawei.com
Cc: <mingo@redhat.com>
Cc: <tglx@linutronix.de>
Cc: <bp@alien8.de>
Cc: <dave.hansen@linux.intel.com>
Cc: <x86@kernel.org>
Cc: <hpa@zytor.com>
Cc: <mhiramat@kernel.org>
Cc: <peterz@infradead.org>
Cc: <ast@kernel.org>
Signed-off-by: Gaosheng Cui <cuigaosheng1@huawei.com>
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
|
|
All uses of modifying_ftrace_code have been removed by
commit 768ae4406a5c ("x86/ftrace: Use text_poke()"),
so remove the declaration, too.
Link: https://lkml.kernel.org/r/20220914110437.1436353-2-cuigaosheng1@huawei.com
Cc: <mingo@redhat.com>
Cc: <tglx@linutronix.de>
Cc: <bp@alien8.de>
Cc: <dave.hansen@linux.intel.com>
Cc: <x86@kernel.org>
Cc: <hpa@zytor.com>
Cc: <mhiramat@kernel.org>
Cc: <peterz@infradead.org>
Cc: <ast@kernel.org>
Signed-off-by: Gaosheng Cui <cuigaosheng1@huawei.com>
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
|
|
With CONFIG_CFI_CLANG, the compiler injects a type preamble immediately
before each function and a check to validate the target function type
before indirect calls:
; type preamble
__cfi_function:
mov <id>, %eax
function:
...
; indirect call check
mov -<id>,%r10d
add -0x4(%r11),%r10d
je .Ltmp1
ud2
.Ltmp1:
call __x86_indirect_thunk_r11
Add error handling code for the ud2 traps emitted for the checks, and
allow CONFIG_CFI_CLANG to be selected on x86_64.
This produces the following oops on CFI failure (generated using lkdtm):
[ 21.441706] CFI failure at lkdtm_indirect_call+0x16/0x20 [lkdtm]
(target: lkdtm_increment_int+0x0/0x10 [lkdtm]; expected type: 0x7e0c52a)
[ 21.444579] invalid opcode: 0000 [#1] PREEMPT SMP NOPTI
[ 21.445296] CPU: 0 PID: 132 Comm: sh Not tainted
5.19.0-rc8-00020-g9f27360e674c #1
[ 21.445296] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996),
BIOS rel-1.14.0-0-g155821a1990b-prebuilt.qemu.org 04/01/2014
[ 21.445296] RIP: 0010:lkdtm_indirect_call+0x16/0x20 [lkdtm]
[ 21.445296] Code: 52 1c c0 48 c7 c1 c5 50 1c c0 e9 25 48 2a cc 0f 1f
44 00 00 49 89 fb 48 c7 c7 50 b4 1c c0 41 ba 5b ad f3 81 45 03 53 f8
[ 21.445296] RSP: 0018:ffffa9f9c02ffdc0 EFLAGS: 00000292
[ 21.445296] RAX: 0000000000000027 RBX: ffffffffc01cb300 RCX: 385cbbd2e070a700
[ 21.445296] RDX: 0000000000000000 RSI: c0000000ffffdfff RDI: ffffffffc01cb450
[ 21.445296] RBP: 0000000000000006 R08: 0000000000000000 R09: ffffffff8d081610
[ 21.445296] R10: 00000000bcc90825 R11: ffffffffc01c2fc0 R12: 0000000000000000
[ 21.445296] R13: ffffa31b827a6000 R14: 0000000000000000 R15: 0000000000000002
[ 21.445296] FS: 00007f08b42216a0(0000) GS:ffffa31b9f400000(0000)
knlGS:0000000000000000
[ 21.445296] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 21.445296] CR2: 0000000000c76678 CR3: 0000000001940000 CR4: 00000000000006f0
[ 21.445296] Call Trace:
[ 21.445296] <TASK>
[ 21.445296] lkdtm_CFI_FORWARD_PROTO+0x30/0x50 [lkdtm]
[ 21.445296] direct_entry+0x12d/0x140 [lkdtm]
[ 21.445296] full_proxy_write+0x5d/0xb0
[ 21.445296] vfs_write+0x144/0x460
[ 21.445296] ? __x64_sys_wait4+0x5a/0xc0
[ 21.445296] ksys_write+0x69/0xd0
[ 21.445296] do_syscall_64+0x51/0xa0
[ 21.445296] entry_SYSCALL_64_after_hwframe+0x63/0xcd
[ 21.445296] RIP: 0033:0x7f08b41a6fe1
[ 21.445296] Code: be 07 00 00 00 41 89 c0 e8 7e ff ff ff 44 89 c7 89
04 24 e8 91 c6 02 00 8b 04 24 48 83 c4 68 c3 48 63 ff b8 01 00 00 03
[ 21.445296] RSP: 002b:00007ffcdf65c2e8 EFLAGS: 00000246 ORIG_RAX: 0000000000000001
[ 21.445296] RAX: ffffffffffffffda RBX: 00007f08b4221690 RCX: 00007f08b41a6fe1
[ 21.445296] RDX: 0000000000000012 RSI: 0000000000c738f0 RDI: 0000000000000001
[ 21.445296] RBP: 0000000000000001 R08: fefefefefefefeff R09: fefefefeffc5ff4e
[ 21.445296] R10: 00007f08b42222b0 R11: 0000000000000246 R12: 0000000000c738f0
[ 21.445296] R13: 0000000000000012 R14: 00007ffcdf65c401 R15: 0000000000c70450
[ 21.445296] </TASK>
[ 21.445296] Modules linked in: lkdtm
[ 21.445296] Dumping ftrace buffer:
[ 21.445296] (ftrace buffer empty)
[ 21.471442] ---[ end trace 0000000000000000 ]---
[ 21.471811] RIP: 0010:lkdtm_indirect_call+0x16/0x20 [lkdtm]
[ 21.472467] Code: 52 1c c0 48 c7 c1 c5 50 1c c0 e9 25 48 2a cc 0f 1f
44 00 00 49 89 fb 48 c7 c7 50 b4 1c c0 41 ba 5b ad f3 81 45 03 53 f8
[ 21.474400] RSP: 0018:ffffa9f9c02ffdc0 EFLAGS: 00000292
[ 21.474735] RAX: 0000000000000027 RBX: ffffffffc01cb300 RCX: 385cbbd2e070a700
[ 21.475664] RDX: 0000000000000000 RSI: c0000000ffffdfff RDI: ffffffffc01cb450
[ 21.476471] RBP: 0000000000000006 R08: 0000000000000000 R09: ffffffff8d081610
[ 21.477127] R10: 00000000bcc90825 R11: ffffffffc01c2fc0 R12: 0000000000000000
[ 21.477959] R13: ffffa31b827a6000 R14: 0000000000000000 R15: 0000000000000002
[ 21.478657] FS: 00007f08b42216a0(0000) GS:ffffa31b9f400000(0000)
knlGS:0000000000000000
[ 21.479577] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 21.480307] CR2: 0000000000c76678 CR3: 0000000001940000 CR4: 00000000000006f0
[ 21.481460] Kernel panic - not syncing: Fatal exception
Signed-off-by: Sami Tolvanen <samitolvanen@google.com>
Reviewed-by: Kees Cook <keescook@chromium.org>
Tested-by: Kees Cook <keescook@chromium.org>
Tested-by: Nathan Chancellor <nathan@kernel.org>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Kees Cook <keescook@chromium.org>
Link: https://lore.kernel.org/r/20220908215504.3686827-23-samitolvanen@google.com
|
|
Disable CONFIG_CFI_CLANG for the stand-alone purgatory.ro.
Signed-off-by: Sami Tolvanen <samitolvanen@google.com>
Reviewed-by: Nick Desaulniers <ndesaulniers@google.com>
Reviewed-by: Kees Cook <keescook@chromium.org>
Tested-by: Nick Desaulniers <ndesaulniers@google.com>
Tested-by: Sedat Dilek <sedat.dilek@gmail.com>
Tested-by: Kees Cook <keescook@chromium.org>
Tested-by: Nathan Chancellor <nathan@kernel.org>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Kees Cook <keescook@chromium.org>
Link: https://lore.kernel.org/r/20220908215504.3686827-22-samitolvanen@google.com
|
|
With CONFIG_CFI_CLANG, assembly functions indirectly called
from C code must be annotated with type identifiers to pass CFI
checking. Define the __CFI_TYPE helper macro to match the compiler
generated function preamble, and ensure SYM_TYPED_FUNC_START also
emits ENDBR with IBT.
Signed-off-by: Sami Tolvanen <samitolvanen@google.com>
Reviewed-by: Kees Cook <keescook@chromium.org>
Tested-by: Kees Cook <keescook@chromium.org>
Tested-by: Nathan Chancellor <nathan@kernel.org>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Kees Cook <keescook@chromium.org>
Link: https://lore.kernel.org/r/20220908215504.3686827-21-samitolvanen@google.com
|
|
The compiler generates __kcfi_typeid_ symbols for annotating assembly
functions with type information. These are constants that can be
referenced in assembly code and are resolved by the linker. Ignore
them in relocs.
Signed-off-by: Sami Tolvanen <samitolvanen@google.com>
Reviewed-by: Kees Cook <keescook@chromium.org>
Tested-by: Kees Cook <keescook@chromium.org>
Tested-by: Nathan Chancellor <nathan@kernel.org>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Kees Cook <keescook@chromium.org>
Link: https://lore.kernel.org/r/20220908215504.3686827-20-samitolvanen@google.com
|
|
In preparation for removing CC_FLAGS_CFI from CC_FLAGS_LTO, explicitly
filter out CC_FLAGS_CFI in all the makefiles where we currently filter
out CC_FLAGS_LTO.
Signed-off-by: Sami Tolvanen <samitolvanen@google.com>
Reviewed-by: Kees Cook <keescook@chromium.org>
Tested-by: Kees Cook <keescook@chromium.org>
Tested-by: Nathan Chancellor <nathan@kernel.org>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Kees Cook <keescook@chromium.org>
Link: https://lore.kernel.org/r/20220908215504.3686827-2-samitolvanen@google.com
|
|
KVM_REQ_UNHALT is now unnecessary because it is replaced by the return
value of kvm_vcpu_block/kvm_vcpu_halt. Remove it.
No functional change intended.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Sean Christopherson <seanjc@google.com>
Acked-by: Marc Zyngier <maz@kernel.org>
Message-Id: <20220921003201.1441511-13-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
KVM_REQ_UNHALT is a weird request that simply reports the value of
kvm_arch_vcpu_runnable() on exit from kvm_vcpu_halt(). Only
MIPS and x86 are looking at it, the others just clear it. Check
the state of the vCPU directly so that the request is handled
as a nop on all architectures.
No functional change intended, except for corner cases where an
event arrive immediately after a signal become pending or after
another similar host-side event.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Sean Christopherson <seanjc@google.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <20220921003201.1441511-12-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
kvm_vcpu_check_block() is called while not in TASK_RUNNING, and therefore
it cannot sleep. Writing to guest memory is therefore forbidden, but it
can happen on AMD processors if kvm_check_nested_events() causes a vmexit.
Fortunately, all events that are caught by kvm_check_nested_events() are
also recognized by kvm_vcpu_has_events() through vendor callbacks such as
kvm_x86_interrupt_allowed() or kvm_x86_ops.nested_ops->has_events(), so
remove the call and postpone the actual processing to vcpu_block().
Opportunistically honor the return of kvm_check_nested_events(). KVM
punted on the check in kvm_vcpu_running() because the only error path is
if vmx_complete_nested_posted_interrupt() fails, in which case KVM exits
to userspace with "internal error" i.e. the VM is likely dead anyways so
it wasn't worth overloading the return of kvm_vcpu_running().
Add the check mostly so that KVM is consistent with itself; the return of
the call via kvm_apic_accept_events()=>kvm_check_nested_events() that
immediately follows _is_ checked.
Reported-by: Maxim Levitsky <mlevitsk@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
[sean: check and handle return of kvm_check_nested_events()]
Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20220921003201.1441511-11-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Don't snapshot pending INIT/SIPI events prior to checking nested events,
architecturally there's nothing wrong with KVM processing (dropping) a
SIPI that is received immediately after synthesizing a VM-Exit. Taking
and consuming the snapshot makes the flow way more subtle than it needs
to be, e.g. nVMX consumes/clears events that trigger VM-Exit (INIT/SIPI),
and so at first glance it appears that KVM is double-dipping on pending
INITs and SIPIs. But that's not the case because INIT is blocked
unconditionally in VMX root mode the CPU cannot be in wait-for_SIPI after
VM-Exit, i.e. the paths that truly consume the snapshot are unreachable
if apic->pending_events is modified by kvm_check_nested_events().
nSVM is a similar story as GIF is cleared by the CPU on VM-Exit; INIT is
blocked regardless of whether or not it was pending prior to VM-Exit.
Drop the snapshot logic so that a future fix doesn't create weirdness
when kvm_vcpu_running()'s call to kvm_check_nested_events() is moved to
vcpu_block(). In that case, kvm_check_nested_events() will be called
immediately before kvm_apic_accept_events(), which raises the obvious
question of why that change doesn't break the snapshot logic.
Note, there is a subtle functional change. Previously, KVM would clear
pending SIPIs if and only SIPI was pending prior to VM-Exit, whereas now
KVM clears pending SIPI unconditionally if INIT+SIPI are blocked. The
latter is architecturally allowed, as SIPI is ignored if the CPU is not
in wait-for-SIPI mode (arguably, KVM should be even more aggressive in
dropping SIPIs). It is software's responsibility to ensure the SIPI is
delivered, i.e. software shouldn't be firing INIT-SIPI at a CPU until
it knows with 100% certaining that the target CPU isn't in VMX root mode.
Furthermore, the existing code is extra weird as SIPIs that arrive after
VM-Exit _are_ dropped if there also happened to be a pending SIPI before
VM-Exit.
Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20220921003201.1441511-10-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Explicitly check for a pending INIT/SIPI event when emulating VMXOFF
instead of blindly making an event request. There's obviously no need
to evaluate events if none are pending.
Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20220921003201.1441511-9-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Evaluate interrupts, i.e. set KVM_REQ_EVENT, if INIT or SIPI is pending
when emulating nested VM-Enter. INIT is blocked while the CPU is in VMX
root mode, but not in VMX non-root, i.e. becomes unblocked on VM-Enter.
This bug has been masked by KVM calling ->check_nested_events() in the
core run loop, but that hack will be fixed in the near future.
Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20220921003201.1441511-8-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Set KVM_REQ_EVENT if INIT or SIPI is pending when the guest enables GIF.
INIT in particular is blocked when GIF=0 and needs to be processed when
GIF is toggled to '1'. This bug has been masked by (a) KVM calling
->check_nested_events() in the core run loop and (b) hypervisors toggling
GIF from 0=>1 only when entering guest mode (L1 entering L2).
Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20220921003201.1441511-7-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Do not return true from kvm_vcpu_has_events() if the vCPU isn' going to
immediately process a pending INIT/SIPI. INIT/SIPI shouldn't be treated
as wake events if they are blocked.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
[sean: rebase onto refactored INIT/SIPI helpers, massage changelog]
Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20220921003201.1441511-6-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Rename kvm_apic_has_events() to kvm_apic_has_pending_init_or_sipi() so
that it's more obvious that "events" really just means "INIT or SIPI".
Opportunistically clean up a weirdly worded comment that referenced
kvm_apic_has_events() instead of kvm_apic_accept_events().
No functional change intended.
Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20220921003201.1441511-5-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Rename and invert kvm_vcpu_latch_init() to kvm_apic_init_sipi_allowed()
so as to match the behavior of {interrupt,nmi,smi}_allowed(), and expose
the helper so that it can be used by kvm_vcpu_has_events() to determine
whether or not an INIT or SIPI is pending _and_ can be taken immediately.
Opportunistically replaced usage of the "latch" terminology with "blocked"
and/or "allowed", again to align with KVM's terminology used for all other
event types.
No functional change intended.
Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20220921003201.1441511-4-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Set KVM_REQ_EVENT when MTF becomes pending to ensure that KVM will run
through inject_pending_event() and thus vmx_check_nested_events() prior
to re-entering the guest.
MTF currently works by virtue of KVM's hack that calls
kvm_check_nested_events() from kvm_vcpu_running(), but that hack will
be removed in the near future. Until that call is removed, the patch
introduces no real functional change.
Fixes: 5ef8acbdd687 ("KVM: nVMX: Emulate MTF when performing instruction emulation")
Cc: stable@vger.kernel.org
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20220921003201.1441511-3-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Interrupts, NMIs etc. sent while in guest mode are already handled
properly by the *_interrupt_allowed callbacks, but other events can
cause a vCPU to be runnable that are specific to guest mode.
In the case of VMX there are two, the preemption timer and the
monitor trap. The VMX preemption timer is already special cased via
the hv_timer_pending callback, but the purpose of the callback can be
easily extended to MTF or in fact any other event that can occur only
in guest mode.
Rename the callback and add an MTF check; kvm_arch_vcpu_runnable()
now can return true if an MTF is pending, without relying on
kvm_vcpu_running()'s call to kvm_check_nested_events(). Until that call
is removed, however, the patch introduces no functional change.
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20220921003201.1441511-2-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Allow force_emulation_prefix to be written by privileged userspace
without reloading KVM. The param does not have any persistent affects
and is trivial to snapshot.
Signed-off-by: Sean Christopherson <seanjc@google.com>
Link: https://lore.kernel.org/r/20220830231614.3580124-28-seanjc@google.com
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Rename inject_pending_events() to kvm_check_and_inject_events() in order
to capture the fact that it handles more than just pending events, and to
(mostly) align with kvm_check_nested_events(), which omits the "inject"
for brevity.
Add a comment above kvm_check_and_inject_events() to provide a high-level
synopsis, and to document a virtualization hole (KVM erratum if you will)
that exists due to KVM not strictly tracking instruction boundaries with
respect to coincident instruction restarts and asynchronous events.
No functional change inteded.
Signed-off-by: Sean Christopherson <seanjc@google.com>
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Link: https://lore.kernel.org/r/20220830231614.3580124-25-seanjc@google.com
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Document the oddities of ICEBP interception (trap-like #DB is intercepted
as a fault-like exception), and how using VMX's inner "skip" helper
deliberately bypasses the pending MTF and single-step #DB logic.
No functional change intended.
Signed-off-by: Sean Christopherson <seanjc@google.com>
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Link: https://lore.kernel.org/r/20220830231614.3580124-24-seanjc@google.com
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Treat pending TRIPLE_FAULTS as pending exceptions. A triple fault is an
exception for all intents and purposes, it's just not tracked as such
because there's no vector associated the exception. E.g. if userspace
were to set vcpu->request_interrupt_window while running L2 and L2 hit a
triple fault, a triple fault nested VM-Exit should be synthesized to L1
before exiting to userspace with KVM_EXIT_IRQ_WINDOW_OPEN.
Link: https://lore.kernel.org/all/YoVHAIGcFgJit1qp@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Link: https://lore.kernel.org/r/20220830231614.3580124-23-seanjc@google.com
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Morph pending exceptions to pending VM-Exits (due to interception) when
the exception is queued instead of waiting until nested events are
checked at VM-Entry. This fixes a longstanding bug where KVM fails to
handle an exception that occurs during delivery of a previous exception,
KVM (L0) and L1 both want to intercept the exception (e.g. #PF for shadow
paging), and KVM determines that the exception is in the guest's domain,
i.e. queues the new exception for L2. Deferring the interception check
causes KVM to esclate various combinations of injected+pending exceptions
to double fault (#DF) without consulting L1's interception desires, and
ends up injecting a spurious #DF into L2.
KVM has fudged around the issue for #PF by special casing emulated #PF
injection for shadow paging, but the underlying issue is not unique to
shadow paging in L0, e.g. if KVM is intercepting #PF because the guest
has a smaller maxphyaddr and L1 (but not L0) is using shadow paging.
Other exceptions are affected as well, e.g. if KVM is intercepting #GP
for one of SVM's workaround or for the VMware backdoor emulation stuff.
The other cases have gone unnoticed because the #DF is spurious if and
only if L1 resolves the exception, e.g. KVM's goofs go unnoticed if L1
would have injected #DF anyways.
The hack-a-fix has also led to ugly code, e.g. bailing from the emulator
if #PF injection forced a nested VM-Exit and the emulator finds itself
back in L1. Allowing for direct-to-VM-Exit queueing also neatly solves
the async #PF in L2 mess; no need to set a magic flag and token, simply
queue a #PF nested VM-Exit.
Deal with event migration by flagging that a pending exception was queued
by userspace and check for interception at the next KVM_RUN, e.g. so that
KVM does the right thing regardless of the order in which userspace
restores nested state vs. event state.
When "getting" events from userspace, simply drop any pending excpetion
that is destined to be intercepted if there is also an injected exception
to be migrated. Ideally, KVM would migrate both events, but that would
require new ABI, and practically speaking losing the event is unlikely to
be noticed, let alone fatal. The injected exception is captured, RIP
still points at the original faulting instruction, etc... So either the
injection on the target will trigger the same intercepted exception, or
the source of the intercepted exception was transient and/or
non-deterministic, thus dropping it is ok-ish.
Fixes: a04aead144fd ("KVM: nSVM: fix running nested guests when npt=0")
Fixes: feaf0c7dc473 ("KVM: nVMX: Do not generate #DF if #PF happens during exception delivery into L2")
Cc: Jim Mattson <jmattson@google.com>
Signed-off-by: Sean Christopherson <seanjc@google.com>
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Link: https://lore.kernel.org/r/20220830231614.3580124-22-seanjc@google.com
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Add a gigantic comment above vmx_check_nested_events() to document the
priorities of all known events on Intel CPUs. Intel's SDM doesn't
include VMX-specific events in its "Priority Among Concurrent Events",
which makes it painfully difficult to suss out the correct priority
between things like Monitor Trap Flag VM-Exits and pending #DBs.
Kudos to Jim Mattson for doing the hard work of collecting and
interpreting the priorities from various locations throughtout the SDM
(because putting them all in one place in the SDM would be too easy).
Cc: Jim Mattson <jmattson@google.com>
Signed-off-by: Sean Christopherson <seanjc@google.com>
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Link: https://lore.kernel.org/r/20220830231614.3580124-21-seanjc@google.com
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Add a helper to identify "low"-priority #DB traps, i.e. trap-like #DBs
that aren't TSS T flag #DBs, and tweak the related code to operate on any
queued exception. A future commit will separate exceptions that are
intercepted by L1, i.e. cause nested VM-Exit, from those that do NOT
trigger nested VM-Exit. I.e. there will be multiple exception structs
and multiple invocations of the helpers.
No functional change intended.
Signed-off-by: Sean Christopherson <seanjc@google.com>
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Link: https://lore.kernel.org/r/20220830231614.3580124-20-seanjc@google.com
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Determine whether or not new events can be injected after checking nested
events. If a VM-Exit occurred during nested event handling, any previous
event that needed re-injection is gone from's KVM perspective; the event
is captured in the vmc*12 VM-Exit information, but doesn't exist in terms
of what needs to be done for entry to L1.
Signed-off-by: Sean Christopherson <seanjc@google.com>
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Link: https://lore.kernel.org/r/20220830231614.3580124-19-seanjc@google.com
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Perform nested event checks before re-injecting exceptions/events into
L2. If a pending exception causes VM-Exit to L1, re-injecting events
into vmcs02 is premature and wasted effort. Take care to ensure events
that need to be re-injected are still re-injected if checking for nested
events "fails", i.e. if KVM needs to force an immediate entry+exit to
complete the to-be-re-injecteed event.
Keep the "can_inject" logic the same for now; it too can be pushed below
the nested checks, but is a slightly riskier change (see past bugs about
events not being properly purged on nested VM-Exit).
Add and/or modify comments to better document the various interactions.
Of note is the comment regarding "blocking" previously injected NMIs and
IRQs if an exception is pending. The old comment isn't wrong strictly
speaking, but it failed to capture the reason why the logic even exists.
Signed-off-by: Sean Christopherson <seanjc@google.com>
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Link: https://lore.kernel.org/r/20220830231614.3580124-18-seanjc@google.com
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Queue #DF by recursing on kvm_multiple_exception() by way of
kvm_queue_exception_e() instead of open coding the behavior. This will
allow KVM to Just Work when a future commit moves exception interception
checks (for L2 => L1) into kvm_multiple_exception().
No functional change intended.
Signed-off-by: Sean Christopherson <seanjc@google.com>
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Link: https://lore.kernel.org/r/20220830231614.3580124-17-seanjc@google.com
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Capture nested_run_pending as block_pending_exceptions so that the logic
of why exceptions are blocked only needs to be documented once instead of
at every place that employs the logic.
No functional change intended.
Signed-off-by: Sean Christopherson <seanjc@google.com>
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Link: https://lore.kernel.org/r/20220830231614.3580124-16-seanjc@google.com
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Move the definition of "struct kvm_queued_exception" out of kvm_vcpu_arch
in anticipation of adding a second instance in kvm_vcpu_arch to handle
exceptions that occur when vectoring an injected exception and are
morphed to VM-Exit instead of leading to #DF.
Opportunistically take advantage of the churn to rename "nr" to "vector".
No functional change intended.
Signed-off-by: Sean Christopherson <seanjc@google.com>
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Link: https://lore.kernel.org/r/20220830231614.3580124-15-seanjc@google.com
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Rename the kvm_x86_ops hook for exception injection to better reflect
reality, and to align with pretty much every other related function name
in KVM.
No functional change intended.
Signed-off-by: Sean Christopherson <seanjc@google.com>
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Link: https://lore.kernel.org/r/20220830231614.3580124-14-seanjc@google.com
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Treat #PFs that occur during emulation of ENCLS as, wait for it, emulated
page faults. Practically speaking, this is a glorified nop as the
exception is never of the nested flavor, and it's extremely unlikely the
guest is relying on the side effect of an implicit INVLPG on the faulting
address.
Fixes: 70210c044b4e ("KVM: VMX: Add SGX ENCLS[ECREATE] handler to enforce CPUID restrictions")
Signed-off-by: Sean Christopherson <seanjc@google.com>
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Link: https://lore.kernel.org/r/20220830231614.3580124-13-seanjc@google.com
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Clear mtf_pending on nested VM-Exit instead of handling the clear on a
case-by-case basis in vmx_check_nested_events(). The pending MTF should
never survive nested VM-Exit, as it is a property of KVM's run of the
current L2, i.e. should never affect the next L2 run by L1. In practice,
this is likely a nop as getting to L1 with nested_run_pending is
impossible, and KVM doesn't correctly handle morphing a pending exception
that occurs on a prior injected exception (need for re-injected exception
being the other case where MTF isn't cleared). However, KVM will
hopefully soon correctly deal with a pending exception on top of an
injected exception.
Add a TODO to document that KVM has an inversion priority bug between
SMIs and MTF (and trap-like #DBS), and that KVM also doesn't properly
save/restore MTF across SMI/RSM.
Signed-off-by: Sean Christopherson <seanjc@google.com>
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Link: https://lore.kernel.org/r/20220830231614.3580124-12-seanjc@google.com
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Fall through to handling other pending exception/events for L2 if SIPI
is pending while the CPU is not in Wait-for-SIPI. KVM correctly ignores
the event, but incorrectly returns immediately, e.g. a SIPI coincident
with another event could lead to KVM incorrectly routing the event to L1
instead of L2.
Fixes: bf0cd88ce363 ("KVM: x86: emulate wait-for-SIPI and SIPI-VMExit")
Signed-off-by: Sean Christopherson <seanjc@google.com>
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Link: https://lore.kernel.org/r/20220830231614.3580124-11-seanjc@google.com
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Use DR7_GD in the emulator instead of open coding the check, and drop a
comically wrong comment.
Signed-off-by: Sean Christopherson <seanjc@google.com>
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Link: https://lore.kernel.org/r/20220830231614.3580124-10-seanjc@google.com
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Add a dedicated "exception type" for #DBs, as #DBs can be fault-like or
trap-like depending the sub-type of #DB, and effectively defer the
decision of what to do with the #DB to the caller.
For the emulator's two calls to exception_type(), treat the #DB as
fault-like, as the emulator handles only code breakpoint and general
detect #DBs, both of which are fault-like.
For event injection, which uses exception_type() to determine whether to
set EFLAGS.RF=1 on the stack, keep the current behavior of not setting
RF=1 for #DBs. Intel and AMD explicitly state RF isn't set on code #DBs,
so exempting by failing the "== EXCPT_FAULT" check is correct. The only
other fault-like #DB is General Detect, and despite Intel and AMD both
strongly implying (through omission) that General Detect #DBs should set
RF=1, hardware (multiple generations of both Intel and AMD), in fact does
not. Through insider knowledge, extreme foresight, sheer dumb luck, or
some combination thereof, KVM correctly handled RF for General Detect #DBs.
Fixes: 38827dbd3fb8 ("KVM: x86: Do not update EFLAGS on faulting emulation")
Cc: stable@vger.kernel.org
Signed-off-by: Sean Christopherson <seanjc@google.com>
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Link: https://lore.kernel.org/r/20220830231614.3580124-9-seanjc@google.com
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Service TSS T-flag #DBs prior to pending MTFs, as such #DBs are higher
priority than MTF. KVM itself doesn't emulate TSS #DBs, and any such
exceptions injected from L1 will be handled by hardware (or morphed to
a fault-like exception if injection fails), but theoretically userspace
could pend a TSS T-flag #DB in conjunction with a pending MTF.
Note, there's no known use case this fixes, it's purely to be technically
correct with respect to Intel's SDM.
Cc: Oliver Upton <oupton@google.com>
Cc: Peter Shier <pshier@google.com>
Fixes: 5ef8acbdd687 ("KVM: nVMX: Emulate MTF when performing instruction emulation")
Signed-off-by: Sean Christopherson <seanjc@google.com>
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Link: https://lore.kernel.org/r/20220830231614.3580124-8-seanjc@google.com
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Exclude General Detect #DBs, which have fault-like behavior but also have
a non-zero payload (DR6.BD=1), from nVMX's handling of pending debug
traps. Opportunistically rewrite the comment to better document what is
being checked, i.e. "has a non-zero payload" vs. "has a payload", and to
call out the many caveats surrounding #DBs that KVM dodges one way or
another.
Cc: Oliver Upton <oupton@google.com>
Cc: Peter Shier <pshier@google.com>
Fixes: 684c0422da71 ("KVM: nVMX: Handle pending #DB when injecting INIT VM-exit")
Signed-off-by: Sean Christopherson <seanjc@google.com>
Reviewed-by: Maxim Levitsky <mlevitsk@redhat.com>
Link: https://lore.kernel.org/r/20220830231614.3580124-7-seanjc@google.com
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Suppress code breakpoints if MOV/POP SS blocking is active and the guest
CPU is Intel, i.e. if the guest thinks it's running on an Intel CPU.
Intel CPUs inhibit code #DBs when MOV/POP SS blocking is active, whereas
AMD (and its descendents) do not.
Signed-off-by: Sean Christopherson <seanjc@google.com>
Link: https://lore.kernel.org/r/20220830231614.3580124-6-seanjc@google.com
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
|
Extend force_emulation_prefix to an 'int' and use bit 1 as a flag to
indicate that KVM should clear RFLAGS.RF before emulating, e.g. to allow
tests to force emulation of code breakpoints in conjunction with MOV/POP
SS blocking, which is impossible without KVM intervention as VMX
unconditionally sets RFLAGS.RF on intercepted #UD.
Make the behavior controllable so that tests can also test RFLAGS.RF=1
(again in conjunction with code #DBs).
Note, clearing RFLAGS.RF won't create an infinite #DB loop as the guest's
IRET from the #DB handler will return to the instruction and not the
prefix, i.e. the restart won't force emulation.
Opportunistically convert the permissions to the preferred octal format.
Signed-off-by: Sean Christopherson <seanjc@google.com>
Link: https://lore.kernel.org/r/20220830231614.3580124-5-seanjc@google.com
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
|
Yu Zhao
Peter Xu
Linus Torvalds
Kees Cook
Bill Wendling
Gaosheng Cui
Sami Tolvanen
Paolo Bonzini
Sean Christopherson