I was re-reading this comment and thought I'd gotten my interpretation
backward, but it's actually just very nuanced. Elaborate a bit so I
hopefully don't reinterpret this again in a few more months.
Bug: None
Test: None
Change-Id: I8ca444f2fb143c46e6068f349e9f5eb574fc4b31
We've talked about this many times in the past, but partners struggle to
understand "expected 38, got 22" in these contexts, and I always have to
go and check the header files just to be sure I'm sure.
I actually think the glibc geterrorname_np() function (which would
return "ENOSYS" rather than "Function not implemented") would be more
helpful, but I'll have to go and implement that first, and then come
back.
Being forced to go through all our errno assertions did also make me
want to use a more consistent style for our ENOSYS assertions in
particular --- there's a particularly readable idiom, and I'll also come
back and move more of those checks to the most readable idiom.
I've added a few missing `errno = 0`s before tests, and removed a few
stray `errno = 0`s from tests that don't actually make assertions about
errno, since I had to look at every single reference to errno anyway.
Test: treehugger
Change-Id: Iba7c56f2adc30288c3e00ade106635e515e88179
Same as in uchar_test.cpp: glibc implements a much older unicode
standard which allows these.
Bug: None
Test: this is a test
Change-Id: Iead5eb01d391be85a7b1a034ea9e7f8828e81cdb
Same as the previous change for the uchar tests: the tests are wrong
to match bionic's wrong implementation. Fix the test to encode the bug
for now while I get the tests into good shape, then I'll be back to
fix the bugs and remove the test differences.
Bug: None
Test: this is a test
Change-Id: I1123660994f755f8bac1f2656f6890d5a43310b3
glibc immediately dereferences the locale passed to all wcsto*_l
functions, even if it won't be used, and even if it's
LC_GLOBAL_LOCALE, which isn't a pointer to valid memory.
Bug: None
Test: this is the test
Change-Id: Ia307cbd6a5c5b4b904c978a03e6d06c1cef6ceed
Coming to C23 via WG14 N2630.
This one is a little interesting, because it actually changes existing
behavior. Previously "0b101" would be parsed as "0", "b", "101" by these
functions. I'm led to believe that glibc plans to actually have separate
versions of these functions for C23 and pre-C23, so callers can have the
behavior they (implicitly) specify by virtue of which -std= they compile
with. Android has never really done anything like that, and I'm pretty
sure app developers have more than enough to worry about with API levels
without having to deal with the cartesian product of API level and C
standard.
Therefore, my plan A is "if you're running on Android >= U, you get C23
behavior". My plan B in the (I think unlikely) event that that actually
causes trouble for anyone is "if you're _targeting_ Android >= U, you
get C23 behavior". I don't think we'd actually want to have two versions
of each of these functions under any circumstances --- that seems by far
the most confusing option.
Test: treehugger
Change-Id: I0bbb30315d3fabd306905ad1484361f5d8745935
Modify bionic unit tests that are built for glibc so that they also
build against musl. They don't all pass though:
With glibc:
2 SLOW TESTS
4 TIMEOUT TESTS
313 FAILED TESTS
YOU HAVE 2 DISABLED TESTS
With musl:
11 SLOW TESTS
11 TIMEOUT TESTS
363 FAILED TESTS
YOU HAVE 2 DISABLED TESTS
Bug: 190084016
Test: m bionic-unit-tests-glibc with musl
Test: atest bionic-unit-tests-static
Test: atest --host bionic-unit-tests-glibc with glibc
Change-Id: I79b6eab04fed3cc4392450df5eef2579412edfe1
This commit fixes wchar.wcstold_hex_floats on arm64. On AArch64
(ARM64), the `long double` type has 128 bits and is more precise then
`double` type (64-bit). As a result, `1e100L` is slightly different
from `static_cast<long double>(1e100)`.
This commit fixes the regression by adding 'L' after the floating point
literals. This should work because casting from a higher precision
to lower precision won't lose any precisions.
Test: adb shell /data/nativetest64/bionic-unit-tests/bionic-unit-tests \
--gtest-filter=wchar.wcstold_hex_floats
Test: adb shell /data/nativetest/bionic-unit-tests/bionic-unit-tests
Test: adb shell /data/nativetest64/bionic-unit-tests/bionic-unit-tests
Change-Id: Ibd7b6a5d46c38338b4ca56838d9d272c710b32f6
...by inlining them.
Also fix a couple of harmless bugs in passing. I've added tests, but in
both cases I don't think it was actually possible to hit the bad behavior:
we'd hit another test and fail immediately after in an externally
indistinguishable way.
Bug: N/A
Test: readelf
Change-Id: I8466050b0bfe2b7b94c76b383cf10c1d9d28debd
The parsefloat routines -- which let us pass NaNs and infinities on to
strto(f|d|ld) -- come from NetBSD.
Also fix LP64's strtold to return a NaN, and fix all the architectures
to return quiet NaNs.
Also fix wcstof/wcstod/wcstold to use parsefloat so they support hex
floats.
Lots of new tests.
Bug: http://b/31101647
Change-Id: Id7d46ac2d8acb8770b5e8c445e87cfabfde6f111
Bionic never had this bug, but since the proposed fix is to remove the NDK's
broken code, we should add a regression test here.
Bug: https://code.google.com/p/android/issues/detail?id=80199
Change-Id: I4de21b5da9913cef990bc4d05a7e27562a71a02b
A mistake I made while cleaning this up the first time through.
mbstrtowcs(3) sets the src param to null if it finishes the string.
Change-Id: I6263646e25d9537043b7025fd1dd6ae195f365e2
The len parameter is a _maximum_ length. The previous code was treating
it as an exact length, causing the following typical call to fail:
mbsrtowcs(out, &in, sizeof(out), state); // sizeof(out) > strlen(in)
Change-Id: I48e474fd54ea5f122bc168a4d74bfe08704f28cc
The bug here turned out to be that we hadn't increased the constant
corresponding to the maximum number of bytes in a character to match
our new implementation, so any character requiring more than a byte
in UTF-8 would break our printf family.
Bug: 15439554
Change-Id: I693e5e6eb11c640b5886e848502908ec5fff53b1
Add optimized versions of bcopy and wmemmove for AArch64 based on the
memmove implementation
Change-Id: I82fbe8a7221ce224c567ffcfed7a94a53640fca8
Signed-off-by: Bernhard Rosenkraenzer <Bernhard.Rosenkranzer@linaro.org>
This reverts commit 8167dd7cb9.
For some reason I thought the bcopy change was bzero. The bcopy code doesn't pass our tests, so reverting until I can figure out what's wrong.
Change-Id: Id89fe959ea5105cd58dff6bba8d91a30cc4bcb07
Add optimized versions of bcopy and wmemmove for AArch64 based on the
memmove implementation
Change-Id: Ie43d0ff4f8ec4edba5b4fb5ccacd941f81ac6557
Signed-off-by: Bernhard Rosenkraenzer <Bernhard.Rosenkranzer@linaro.org>
I've also switched some tests to be positive rather than negative,
because !defined is slightly harder to reason about and there are
only two cases: bionic and glibc.
Change-Id: I8d3ac40420ca5aead3e88c69cf293f267273c8ef
when compile the cts package with aarch64 gcc4.9, will get following error:
bionic/tests/wchar_test.cpp:253:3: required from here
external/gtest/include/gtest/gtest.h:1448:16:
error: comparison between signed and unsigned integer expressions [-Werror=sign-compare]
this change fix it by using static_cast<wchar_t> as suggested by Calin Juravle
Change-Id: I7fb9506e7b84b8a12b9d003458d4f0e78554c3cd
Signed-off-by: Yongqin Liu <yongqin.liu@linaro.org>
Although glibc gets by with an 8-byte mbstate_t, OpenBSD uses 12 bytes (of
the 128 bytes it reserves!).
We can actually implement UTF-8 encoding/decoding with a 0-byte mbstate_t
which means we can make things work on LP32 too, as long as we accept the
limitation that the caller needs to present us with a complete sequence
before we'll process it.
Our behavior is fine when going from characters to bytes; we just
update the source wchar_t** to say how far through the input we got.
I'll come back and use the 4 bytes we do have to cope with byte sequences
split across multiple input buffers. The fact that we don't support
UTF-8 sequences longer than 4 bytes plus the fact that the first byte of
a UTF-8 sequence encodes the length means we shouldn't need the other
fields OpenBSD used (at the cost of some recomputation in cases where a
sequence is split across buffers).
This patch also makes the minimal changes necessary to setlocale(3) to
make us behave like glibc when an app requests UTF-8. (The difference
being that our "C" locale is the same as our "C.UTF-8" locale.)
Change-Id: Ied327a8c4643744b3611bf6bb005a9b389ba4c2f
This replaces a partial set of non-functional functions with a complete
set of functions, all of which actually work.
This requires us to implement mbsnrtowcs and wcsnrtombs which completes
the set of what we need for libc++.
The mbsnrtowcs is basically a copy & paste of wcsnrtombs, but I'm going
to go straight to looking at using the OpenBSD UTF-8 implementation rather
than keep polishing our home-grown turd.
(This patch also opportunistically switches us over to upstream btowc,
mbrlen, and wctob, since they're all trivially expressed in terms of
other functions.)
Change-Id: I0f81443840de0f1aa73b96f0b51988976793a323
