The memcpy implementation is not completely stupid but is almost
certainly not as good as a carefully tuned and optimised one.
Building snmalloc with FreeBSD's libc memcpy + jemalloc and with this,
each 10 times, does not show a statistically significant performance
difference at 95% confidence. The snmalloc version has very slightly
lower median and worst-case times. This is in no way a sensible
benchmark, but it serves as a smoke test for significant performance
regressions.
The CI self-host job now uses the checked memcpy.
This also fixes an off-by-one error in the external bounds. This is
triggered by ninja, so we will see breakage in CI if it is reintroduced.
In debug builds, we provide a verbose error containing the address of
the allocation, the base and bounds of the allocation, and a backtrace.
The backtrace was broken by the CI cleanup moving the BACKTRACE_HEADER
macro into the SNMALLOC_ namespace. This is also fixed.
The test involves hijacking `abort`, which doesn't work everywhere. It
also requires `backtrace` to work in configurations where stack traces
are enabled. This is disabled in QEMU because `backtrace` appears to
crash reliably in QEMU user mode.
For now, in the -checks build configurations, we are hitting a slow path
in the pagemap on accesses so that the pages that are `PROT_NONE` don't
cause crashes. These need to be made read-only, but this requires a PAL
change.
Introduce Metaslab::from_link(SlabLink*) to encapsulate the "container of"
dance. Note that Metaslab was not a standard layout type prior to this change
(since both SlabLink and Metaslab defined non-static data members), and so the
reinterpret_cast<>s replaced here with ::from_link() were UB, but everyone lays
out classes as one expects so it was fine in practice.
Most of the uses of ::from_link() are already guarded by checks that the link
pointer is not nullptr, but in src/mem/corealloc.h:/debug_is_empty_impl we shift
to testing the link pointer explicitly before converting to the metaslab.
Despite that Metaslab is now standard layout, we still don't fall back to the
inter-convertibility of a standard layout class and its first[*] data member
since we're going to want to put a common initial sequence across Metaslab and
ChunkRecord and the SlabLink isn't likely to be in it.
Modernise and tidy the CMake a bit:
- Use generator expressions for a lot of conditionals so that things
are more reliable with multi-config generators (and less verbose).
- Remove C as a needed language. None of the code was C but we were
using C to test if headers worked. This was fragile because a build
with `CMAKE_CXX_COMPILER` set might have checked things compiled with
the system C compiler and then failed when the specified C++ compiler
used different headers.
- Rename the `BACKTRACE_HEADER` macro to `SNMALLOC_BACKTRACE_HEADER`.
This is exposed into code that consumes snmalloc and so should be
'namespaced' (to the degree that's possible with C macros).
- Clean up the options and use dependent options to hide options
that are not always relevant.
- Use functions instead of macros for better variable scoping.
- Factor out some duplicated bits into functions.
- Update to the latest way of telling CMake to use C++17 or C++20.
- Migrate everything that's setting global properties to setting only
per-target properties.
- Link with -nostdlib++ if it's available. If it isn't, fall back to
enabling the C language and linking with the C compiler.
- Make the per-test log messages verbose outputs. These kept scrolling
important messages off the top of the screen for me.
- Make building as a header-only library a public option.
- Add install targets that install all of the headers and provide a
config option. This works with the header-only configuration for
integration with things like vcpkg.
- Fix a missing `#endif` in the `malloc_useable_size` check. This was
failing co compile on all platforms because of the missing `#endif`.
- Bump the minimum version to 3.14 so that we have access to
target_link_options. This is necessary to use generator expressions
for linker flags.
- Make the linker error if the shim libraries depend on symbols that
are not defined in the explicitly-provided libraries.
- Make the old-Ubuntu CI jobs use C++17 explicitly (previously CMake
was silently ignoring the fact that the compiler didn't support C++20)
- Fix errors found by the more aggressive linking mode.
With these changes, it's now possible to install snmalloc and then, in
another project, do something like this:
```cmake
find_package(snmalloc CONFIG REQUIRED)
target_link_libraries(t1 snmalloc::snmalloc)
target_link_libraries(t2 snmalloc::snmallocshim-static)
```
In this example, `t1` gets all of the compile flags necessary to include
snmalloc headers for its build configuration. `t2` is additionally
linked to the snmalloc static shim library.
David points out that we might not have a static way to get at the pagemap, so
it is potentially useful to pass pointers to state objects down from the
Allocators.
Fortunately, C++ taketh away and C++ giveth, both, so here we are: a way to
detect if we're in the middle of definining a type that uses itself as a
template parameter in a way that flows into a concept check and, if so,
short-circuit out of the need to actually do any checks. Wonders never cease.
And do so by type, rather than by value. While here, introduce a C++20 concept
for this Backend-offered proxy and adjust the template parameters appropriately.
This will be useful for the process sandbox code, which needs to mediate stores
to the pagemap, but can provide a read-only view.
When we are accessing potentially out of range, then we might be
accessing before the pagemap has been initialised. Move the check
into the pagemap for better codegen.
This commit splits the sizeclass meta-data to generate better cache
locality for various lookups for checking for size and start of
sizeclasses.
Also, contains some tidying including removing sizeclasses covering
large range. This is left over from an alternative design for large
classes that is no longer in use.
This passes though to an underlying allocator rather than using
snmalloc. This is required for using ASAN in Verona. Verona takes a
close coupling with snmalloc, but to use with ASAN would require a
more work, so we pass to the system allocator in this case.
The code was able to use pthread destructors rather than C++ thread
local destructors. This removes the dependence on a C++ .so on linux.
However, this is not stable on other platforms such as Apple. Where the
C++ thread local state can be cleared before the pthread destructor
runs.
Instead, tell the iostream to write out hex. This avoids the CHERI compiler
warning that we're turning a provenance-free value to a pointer.
Co-authored-by: Matthew Parkinson <mattpark@microsoft.com>
This avoids the CHERI compiler warning that we're turning a provenance-free
value to a pointer.
Co-authored-by: Matthew Parkinson <mattpark@microsoft.com>
We mean to be allocating MIN_ALLOC_SIZE (== 2 * sizeof(void*)), not
sizeof(MIN_ALLOC_SIZE) (== sizeof(size_t)). This doesn't matter in practice
since, well, MIN_ALLOC_SIZE is the minimum allocation size, and so requesting
either will have the same effect. Still, best to say what we mean.
With snmalloc2, slabs are linked through the Metaslab structure directly rather
than in-band in a free allocation, so we no longer need to store a SlabLink in
even the smallest allocation classes.
The threshold for waking is used to ensure that we only use a slab when
it has sufficient space that we won't hit the slow path to soon after
using this slab. In the checked version, this is also used to give some
entropy in layout. Changing this to always be a pertcentage in the
checked case increases the effectiveness of the free list to detect OOB
write.
- CI merge issues:
- The malloc shim libraries are renamed.
- CMake gets very unhappy if you don't enable the C language and
tries to link with the C compiler instead of the C++ compiler if
you do enable it.
- The Ubuntu packages for QEMU install a `binfmt_misc` activator for
PowerPC64 little-endian, but set the page size to 4 KiB. We then
tried to run the tests (which expect 64 KiB pages) and became very
confused when `mmap` returned 4 KiB-aligned memory.
- Test failures:
- Fix all of the issues UBsan found.
- Underflow in `pointer_offset` when used to add negative offsets.
- `CoreAlloc`'s `LocalState` accessed on a null `CoreAlloc` pointer.
- Out of bounds access in the sizeclass list on attempts to access
more memory than fits in the VA space.
-
- There was an integer overflow in `AddressSpace` that could cause it
to try to allocate a zero-sized object, get a null pointer, and
then try to do something with 0 - {size of the real allocation}.
- The malloc tests weren't setting `errno` to 0 before doing
calling `malloc`, which should set `errno` on failure, and then
checking that `errno` was 0.
- Don't call `PAL::error` on PAL allocation failure, return `nullptr`.
The PALs were inconsistent about that and the new code expects to be
able to report address-space exhaustion.
- The malloc checks can behave differently with 0-sized allocations
on different platforms but were very fragile about their
expectations.
- The malloc test didn't report failure for all of the ways that it
could fail and so was spuriously passing on some platforms.
- The perf test for external pointer is currently very slow on
Windows. The number of loops have been reduced and a timeout added
for the Windows CI runs.
- The logic to capture `errno` across calls was using
`decltype(errno)`, which on some platforms where `errno` is a macro
evaluated to `int&` and so they captured a reference rather than
the value and failed to reset `errno`.
- The Apple PAL can set `errno` on `notify_using` if it's called with
memory that was not previously passed to `notify_not_using` but was
not adequately protected against this and so would sometimes cause
`malloc` to set `errno` to `EINVAL`.
