These functions should be declared in `stdlib.h`, not `wchar.h`, as
confusing as it is. Move them to the proper header file and matching
directories in src/ and test/ trees.
This was discovered while testing libc++ build against llvm-libc, which
re-declares functions like mbtowc in std-namespace in `<cstdlib>`
header, and then uses those functions in its locale implementation.
Functions like isalpha / tolower can operate on chars internally. This
allows us to get rid of unnecessary casts and open a way to creating
wchar_t overloads with the same names (e.g. for isalpha), that would
simplify templated code for conversion functions (see
315dfe5865962d8a3d60e21d1fffce5214fe54ef).
Add the int->char converstion to public entrypoints implementation
instead. We also need to introduce bounds check on the input argument
values - these functions' behavior is unspecified if the argument is
neither EOF nor fits in "unsigned char" range, but the tests we've had
verified that they always return false for small negative values. To
preserve this behavior, cover it explicitly.
This is a part of allocator patch since I want to make sure the TLS for
allocators are correctly handled.
TLS dtors are not invoked on exit previously. This departures from major
libc implementations.
This PR fixes the issue by aligning the behavior with bionic.
https://android.googlesource.com/platform/bionic/+/refs/heads/main/libc/bionic/exit.cpp
I believe the finalization order is also consistent with glibc now:
On main thread exiting:
- pthread_key dtors are not called (unless exiting with `pthread_exit`)
- `__cxa` based tls dtors are called
- `::__cxa_atexit` and `::atexit` dtors are called
- `.fini` dtors are called
Summary:
This patch uses the actual allocator interface to implement
`aligned_alloc`. We do this by simply rounding up the amount allocated.
Because of how index calculation works, any offset within an allocated
pointer will still map to the same chunk, so we can just adjust
internally and it will free all the same.
Summary:
Now that we have `malloc` we can implement `realloc` efficiently. This
uses the known chunk sizes to avoid unnecessary allocations. We just
return nullptr for NVPTX. I'd remove the list for the entrypoint but
then the libc++ code would stop working. When someone writes the NVPTX
support this will be trivial.
Summary:
Currently we dispatch the writing mode off of a runtime enum passed in
by the constructor. This causes very unfortunate codegen for the GPU
targets where we get worst-case codegen because of the unused function
pointer for `sprintf`. Instead, this patch moves all of this to a
template so it can be masked out. This results in no dynamic stack and
uses 60 VGPRs instead of 117. It also compiles about 5x as fast.
This reverts commit 1e6e845d49a336e9da7ca6c576ec45c0b419b5f6 because it
changed the 1st parameter of adjust() to be unsigned, but libc itself
calls adjust() with a negative argument in align_backward() in
op_generic.h.
Summary:
This is a blocker on another patch in the OpenMP runtime. The problem is
that NVIDIA truly doesn't handle RPC-based allocations very well. It
cannot reliably update the MMU while a kernel is running and it will
usually deadlock if called from a separate thread due to internal use of
TLS.
This patch just removes the definition of `malloc` and `free` for NVPTX.
The result here is that they will be undefined, which is the cue for the
`nvlink` linker to define them for us. So, as far as `libc` is concerned
it still implements malloc.
Add LIBC_ASSERT statements to FixedVector implementation, and zero out
the memory when the elements are removed to flag out-of-bound access and
dangling pointer/reference access.
This change unmasks the bug in one of FixedVector uses for atexit
handlers: dangling reference use, which was actually led to crashes in
the wild (with prod blockstore implementation). Fix it in this CL.
This reverts commit 93b83642ee34d0092b94776728dad0117c2b72a1.
- Correct riscv32 assumption about alignment (bit of a hack).
- Fix test case where the largest_small and smallest sizes are the
same.
- Fix assertion expressions.
- Fix incorrect small size in freestore_test.
- There may only be one small size for high alignment and small
pointers (riscv32).
- Don't rely on stack alignment in FreeList test.
This reworks the free store implementation in libc's malloc to use a
dlmalloc-style binary trie of circularly linked FIFO free lists. This
data structure can be maintained in logarithmic time, but it still
permits a relatively small implementation compared to other
logarithmic-time ordered maps.
The implementation doesn't do the various bitwise tricks or
optimizations used in actual dlmalloc; it instead optimizes for
(relative) readability and minimum code size. Specific optimization can
be added as necessary given future profiling.
Summary:
I'm going to attempt to move the `rpc.h` header to a separate folder
that we can install and include outside of `libc`. Before doing this I'm
going to try to trim up the file so there's not as many things I need to
copy to make it work. This dependency on `cpp::functional` is a low
hanging fruit. I only did it so that I could overload the argument of
the work function so that passing the id was optional in the lambda,
that's not a *huge* deal and it makes it more explicit I suppose.
Previously, the Quicksort implementation was written in the obvious way:
after each partitioning step, it explicitly recursed twice to sort the
two sublists. Now it compares the two sublists' sizes, and recurses only
to sort the smaller one. To handle the larger list it loops back round
to the top of the function, so as to handle it within the existing stack
frame.
This means that every recursive call is handling a list at most half
that of its caller. So the maximum recursive call depth is O(lg N).
Otherwise, in Quicksort's bad cases where each partition step peels off
a small constant number of array elements, the stack usage could grow
linearly with the array being sorted, i.e. it might be Θ(N).
I tested this code by manually constructing a List Of Doom that causes
this particular quicksort implementation to hit its worst case, and
confirming that it recursed very deeply in the old code and doesn't in
the new code. But I haven't added that list to the test suite, because
the List Of Doom has to be constructed in a way based on every detail of
the quicksort algorithm (pivot choice and partitioning strategy), so it
would silently stop being a useful regression test as soon as any detail
changed.
This patch adds the malloc.h header, declaring Scudo's mallopt
entrypoint when built LLVM_LIBC_INCLUDE_SCUDO, as well as two
constants that can be passed to it (M_PURGE and M_PURGE_ALL).
Due to limitations of the current build system, only the declaration
of mallopt is gated by LLVM_LIBC_INCLUDE_SCUDO, and the two new
constants are defined irrespectively of it. We may need to refine
this in the future.
Note that some allocators other than Scudo may offer a mallopt
implementation too (e.g. man 3 mallopt), albeit with different
supported input values. This patch only supports the specific case of
LLVM_LIBC_INCLUDE_SCUDO.
Summary:
This function can easily be implemented by forwarding it to the host
process. This shows up in a few places that we might want to test the
GPU so it should be provided. Also, I find the idea of the GPU
offloading work to the CPU via `system` very funny.
The call chain to `Mutex:lock` can be polluted by stack protector. For
completely safe, let's postpone the main TLS tearing down to a separate
phase.
fix#108030
Summary:
This provides the `_l` variants for the `stdlib.h` functions. These are
just copies of the same entrypoint and don't do anything with the locale
information.
Summary:
This patch implements 'getenv'. I was torn on how to implement this,
since realistically we only have access to this environment pointer in
the "loader" interface. An alternative would be to use an RPC call every
time, but I think that's overkill for what this will be used for. A
better solution is just to emit a common `DataEnvironment` that contains
all of the host visible resources to initialize. Right now this is the
`env_ptr`, `clock_freq`, and `rpc_client`.
I did this by making the `app.h` interface that Linux uses more general,
could possibly move that into a separate patch, but I figured it's
easier to see with the usage.
This is a non-feature change that enables most of the entrypoints for
aarch64 based runtime builds. It fixes an additional problem that some
compiler-rt targets are not defined at the time of dependency checking
thus leading to false-negatives.
