ff63f7b80f
Convert float_to_int! into a function
2017-09-14 14:59:32 +02:00
a272d753f5
Don't use the Wrapping wrapper in float addition
...
Inspired by a patch by mattico.
2017-09-14 02:08:28 +02:00
72ed4c8bce
Don't use a macro for pow calculation
2017-09-14 01:59:02 +02:00
8313cec597
Extend the Float trait by some constants and supertraits
2017-09-14 01:57:52 +02:00
2c1d7dccc1
Use the newly added ONE constant in float addition
2017-09-14 00:10:16 +02:00
46cfa05650
Introduce a float_impl! macro to avoid duplication
2017-09-13 22:44:56 +02:00
9bdedec38e
Refactor float builtins to use associated consts
2017-09-13 22:09:52 +02:00
c9a261895e
Refactor int builtins to use associated consts
2017-09-13 22:09:31 +02:00
c138713df1
Avoid memcpy references in unoptimized code
2017-07-22 23:10:30 +02:00
064452423a
Don't include custom chkstk on MSVC
...
MSVC includes its own __chkstk so these aren't used.
2017-07-12 22:40:49 +01:00
91eaa85902
Don't mangle probes all the time
2017-07-07 20:35:14 -07:00
f9f6bd0589
Tweak definition of probestack functions
...
It looks like the old `__rust_probestack` routine is incompatible with newer
linux kernels. My best guess for this is that the kernel's auto-growth logic is
failing to trigger, causing what looks like a legitimate segfault to get
delivered. My best guess for why *that's* happening is that the faulting address
is below `%rsp`, whereas previously all faulting stack addresses were above
`%rsp`. The probestack routine does not modify `%rsp` as it's probing the stack,
and presumably newer kernels are interpreting this as a legitimate violation.
This commit tweaks the probestack routine to instead update `%rsp` incrementally
as probing happens. The ABI of the function, however, requires that `%rsp`
isn't changed as part of the function so it's restored at the end to the
previous value.
2017-07-07 10:16:03 -07:00
7ccf8408d0
Add `__rust_probestack` intrinsic
...
Will be required for rust-lang/rust#42816
2017-07-06 06:58:32 -07:00
238647af80
Fix unused imports on iOS
2017-07-03 19:11:34 -07:00
681aaa914d
Use the Rust implementation of udivsi3 on ARM
...
Although compiler-rt presumably has a more optimized implementation written in
assembly, it appears buggy for whatever reason, causing #173 .
For now let's see if integration into rust-lang/rust will work with the
Rust-defined implementation!
2017-07-03 09:48:11 -07:00
b8a662040e
no aeabi_mem* symbols on iOS, weak symbols on thumb, normal symbols elsewhere
2017-06-30 18:06:25 -05:00
75c6ccca71
optimize 32-bit aligned mem{cpy,clr,set} intrinsics for ARM
...
this reduces the execution time of all these routines by 40-70%
2017-06-29 22:40:58 -05:00
0ebbcaede4
Enable the `intrinsics` program on thumb
2017-06-25 10:09:50 -07:00
afe5c717a9
Address review comments
2017-06-24 21:51:34 -07:00
c193113721
Don't test mangled names on thumb
...
We are both the "real compiler-rt" and the "to be tested one".
2017-06-24 12:54:35 -07:00
acaa2f071a
Don't check for references to panics with debug assertions
2017-06-24 11:44:50 -07:00
d051480eff
Don't generate unmangled aeabi with gen-tests
...
The symbols they delgate to also don't exist...
2017-06-24 11:34:10 -07:00
91b1291dc5
Don't try to work with `cargo test`
2017-06-24 10:22:49 -07:00
734ec3d31c
Tweak testing and such:
...
* Don't run `intrinsics` tests on thumb
* Disable `compiler_builtins` attribute on `feature = "gen-tests"`
* Disable mangling on `feature = "gen-tests"` instead of `cfg(test)`
2017-06-24 10:10:04 -07:00
2147753559
Remove usage of unwrap_or_else
2017-06-23 21:31:54 -07:00
4dac0be751
Don't derive Debug for Sign
2017-06-23 21:11:31 -07:00
b42d267065
Don't compile assembly on x86_64 Windows
...
They've all got the wrong ABI...
2017-06-23 21:09:24 -07:00
eabb6fab4a
Shore up docs in the macros module
2017-06-23 16:28:50 -07:00
7de57cd4f9
Handle aeabi aliasing
...
Objects in compiler-rt may have two symbols, so this makes sure that we don't
bring in those objects by accident by defining the aliases ourselves.
2017-06-23 15:55:11 -07:00
a839d53a02
Deal with floatdidf on x86_64
...
Apparently LLVM will lower this down to just an instruction
2017-06-23 13:28:24 -07:00
5172f8c218
Tweak usage of C shims on MSVC
2017-06-23 12:03:33 -07:00
a16ebb0dcb
Use `nm` to weed out panics
2017-06-23 11:52:22 -07:00
5a444d58f2
Remove unused rust file
2017-06-23 11:21:36 -07:00
7c4745061a
Remove executable bit on arm.rs
2017-06-23 11:21:15 -07:00
83d63eaa9b
Convert float intrinsics to the `intrinsics!` macro
2017-06-23 11:05:25 -07:00
93fed264c1
Remove executable bit from conv.rs
2017-06-23 10:44:59 -07:00
4c41b5649f
Fix use_c_shim_if
2017-06-23 07:54:31 -07:00
47ff81325c
Use more intrinsics on msvc
2017-06-23 07:53:34 -07:00
12a0038250
Fix the abi hack on windows
2017-06-22 23:36:50 -07:00
d17042106f
Port udiv to the `intrinsics!` macro
2017-06-22 23:23:36 -07:00
7886ae275b
Port sdiv to traits + `intrinsics!`
...
Enhance `intrinsics!` along the way!
2017-06-22 23:09:28 -07:00
275d1032b5
Port mul intrinsics to traits
...
Also add a few features to the `intrinsics!` macro
2017-06-22 22:36:37 -07:00
eeb44abacf
Remove executable bit from mod.rs
2017-06-22 22:04:40 -07:00
2f2bb32e40
Use traits instead of macros for shift intrinsics
...
This is an attempt to tidy up the definition of intrinsics by making them more
rust-like at the definition site and using traits instead of macros for the
definition. Additionally the helper macro, `intrinsics!`, now fills in a
definition for #[cfg]'d off intrinsics when compiling with C code
2017-06-22 22:03:07 -07:00
b2f72e1eb1
remove unused macros
...
to fix warnings
2017-06-01 07:40:50 -05:00
673a6f9948
Fix overflow bug when creating the absolute value
...
Previously, the tests failed on some platforms due to it.
2017-05-08 04:30:48 +02:00
f90792e1bc
Add i128 <-> float conversion functions
2017-05-06 02:22:54 +02:00
144e54d7b0
Rename float conversion macros to something nicer
2017-05-05 23:31:41 +02:00
19b2efdf36
move intrinsics test from src/bin to examples
...
with this change the libc crate is not required when calling `cargo build -p
compiler-builtins`
2017-04-11 17:09:05 -05:00
d34c51f74a
remove arm tests from the old test suite
2017-04-10 16:35:08 -05:00
b8e8c7f693
no_std compatible tester
2017-04-10 14:38:17 -05:00
cc0d8ba2d1
remove the current test suite
2017-04-10 11:14:05 -05:00
28ac4908f8
Auto merge of #147 - rust-lang-nursery:conv, r=japaric
...
Conversion from&to float<->integer
this is a rebased version of #139
cc @ithinuel
2017-04-08 15:20:35 +00:00
109c33ed0a
test out of range condition using f64
...
to avoid imprecision due to rounding to f32
2017-04-08 10:18:19 -05:00
2faec6a4fb
fix debug assertion in modsi3 / moddi3
...
fixes #151
this fix is very similar to #149
2017-03-13 20:40:44 -05:00
9681358dc5
Auto merge of #149 - rust-lang-nursery:gh148, r=alexcrichton
...
fix debug assertion in divdi3
fixes #148
r? @alexcrichton
2017-03-07 00:38:33 +00:00
c93b32b3e5
fix debug assertion in divdi3
...
fixes #148
2017-03-06 17:46:11 -05:00
4c556dc9e7
fix quickcheck tests
...
the check! macro syntax has changed
2017-03-06 12:17:41 -05:00
c6e06f4157
refactor 'impl Int'
2017-03-06 12:17:10 -05:00
09d2f2f1f5
exclude from tests values out of integer range (and NaN).
2017-03-06 11:56:52 -05:00
d96102d270
implement float/double to (u)int conversion.
2017-03-06 11:56:52 -05:00
293fef5ebe
impl (unsigned/signed) int to single/double precision float conversion based on llvm algorithms.
2017-03-06 11:55:57 -05:00
644a1c9ac9
fix infinite recursion in divmoddi4 / mulodi4
...
on ARMv7-M processors, divmoddi4 was calling mulodi4 and mulodi4 was calling
divmoddi4 leading to infinite recursion. This commit breaks the cycle by using
wrapping multiplication in divmoddi4.
fixes #145
2017-03-06 10:38:19 -05:00
5bb969d8f5
implement float subtraction
...
as a + (-b)
2017-02-08 10:10:40 -05:00
47b45d1d62
adjust the check! macro to accept AAPCS intrinsics
2017-02-07 23:07:51 -05:00
57085be2ea
ARM: keep some non-aeabi symbols around
...
- multi3: there's no aeabi equivalent
- divmod{s,d}i4: these are directly called by __aeabi_{l,i}divmod
- add{s,d}f3: required by the C sub{s,d}f3 implementation
but make sure they also use the AAPCS calling convention
2017-02-07 14:28:02 -05:00
dfa7b161aa
use AAPCS calling convention on all aeabi intrinsics
...
also, on ARM, inline(always) the actual implementation of the intrinsics so we
end with code like this:
```
00000000 <__aeabi_dadd>:
(implementation here)
```
instead of "trampolines" like this:
```
00000000 <__aeabi_dadd>:
(shuffle registers)
(call __adddf3)
00000000 <__adddf3>:
(implementation here)
```
closes #116
2017-02-07 09:41:26 -05:00
37d349015e
Disable quickcheck tests on mips
...
Two reasons:
* the C versions __divti3 and __modti3 are apparently broken,
at least when used in quickcheck. They change their own arguments.
* compiler_rt's support for mips is disabled already on clang [1].
Its desireable to support working "cargo test" on that compiler
as well, and not greet the tester with linker errors.
[1]: http://llvm.org/viewvc/llvm-project?view=revision&revision=224488
2017-02-04 12:23:45 +01:00
f8a4e3fd70
Panic when invalid overflow value is returned
2017-02-04 01:39:19 +01:00
b78e956f08
Add quickcheck tests
2017-02-03 23:46:15 +01:00
417b07b821
Add newly implemented intrinsics to test file
2017-02-03 01:51:46 +01:00
9013dbef02
u128 sdiv intrinsics
2017-02-02 22:13:28 +01:00
2f3c534cfe
u128 udiv intrinsics
2017-02-02 22:11:14 +01:00
8fe50d813c
Macro-ify udivmod
2017-02-02 22:03:42 +01:00
1e27c3f937
i128 shift intrinsics
2017-02-02 22:02:51 +01:00
3055aa21b6
i128 mul intrinsics
2017-02-02 21:36:56 +01:00
b356429374
int module: macro-ify trait impls and add {u,i}128 support
2017-02-02 21:36:02 +01:00
a2e2ec1a18
Add i128 to lib.rs as feature
2017-02-02 21:27:11 +01:00
b6f059d627
fix the build
...
the unused macro_use crate lint has tightened; fix the new warnings
2017-01-26 16:17:23 -05:00
ee0aacf4be
mul.rs: use wrapping_mul not plain
...
rustc in debug mode with a plain multiplication will call @llvm.umul.with.overflow.* which may call the builtin resulting in infinite recursion.
2017-01-02 22:57:38 +13:00
53da94d74f
more #[no_mangle]
2016-12-31 10:04:40 -05:00
3992ed1e7a
CI: use a recent nightly
...
- add #[no_mangle] to the panic_fmt lang item to adjust to changes in
the visibility algorithm
- adjust to changes in the layout of Cargo's target directory
- use a newer Xargo to reduce the build time of the sysroot (only core
is compiled as part of the sysroot now)
2016-12-31 10:04:40 -05:00
9897bfb8a9
Fix memset arguments for MSP430 target.
...
On MSP430 second argument to memset has i16 type instead of i32.
2016-12-19 05:31:20 +03:00
a1caa7ccac
add implementations of memcpy et al
...
behind the "mem" Cargo feature, which used to be named "weak"
fixes #126
2016-12-17 23:06:37 -05:00
3484b2b68e
by default, mark this crate as the #![compiler_builtins] crate
...
as this is how most users will want to use this crate
2016-12-11 16:18:43 -05:00
f68475e080
Use better cfg name for arm-linux tests
2016-11-12 15:52:59 -06:00
77ca63c511
Use correct lib name in qc test failure
2016-11-12 15:30:57 -06:00
655f642d3f
Add float quickcheck
2016-11-12 14:55:28 -06:00
4733417140
Fix a few clippy warnings
2016-10-16 13:53:31 +02:00
5d164a4eda
fix the powerpc64le target
...
by selecting a CPU that's compatible with the instructions that `rustc`
generates
2016-10-14 09:59:21 -05:00
a163ca7ac3
fix more warnings
2016-10-10 20:21:11 -05:00
e05caa1227
fix warnings
2016-10-10 19:45:34 -05:00
8f0847eb94
rename to compiler_builtins, add rust-lang/rust-ish crate attributes
2016-10-10 19:05:21 -05:00
24c0247c6a
add the compiler_builtins feature to src/lib.rs
2016-10-10 16:45:24 -05:00
d2a3c45a29
set crate_name and crate_type in src/lib.rs
2016-10-10 16:43:38 -05:00
96ddd4ed8d
use test::black_box instead of ptr::*_volatile
...
both prevent LLVM from optimizing away the intrinsics but the former
doesn't produce an `intrinsics` binary that segfaults
2016-10-07 18:26:21 -05:00
c82403551d
fix unsafe warnings
2016-10-07 18:19:40 -05:00
fd69beba29
get profile.dev (-debug-assertions) + LTO working
2016-10-07 18:19:40 -05:00
a21fb956f9
insert more `abort()` calls where division by zero may occur
2016-10-07 18:19:40 -05:00
a2ffc799d0
check that we don't have any call to panic in our implementations
2016-10-07 18:19:40 -05:00
534d22e287
panic! -> abort
...
closes #79
2016-10-07 18:19:40 -05:00
6622c49be6
Implement powi_f2
2016-10-03 11:20:42 -05:00
02140cddc9
Initial implementation of multitester
2016-10-02 14:38:40 -05:00
3b8dedd416
Revert "Merge pull request #48 from mattico/add_float_quickcheck"
...
This reverts commit e34a6058dfcab88e6133
2016-09-30 19:12:17 -05:00
010d153966
Add Quickcheck types for float tests
2016-09-30 14:29:52 -05:00
9b1fed8d3f
Cleanup add builtin assignments
2016-09-30 14:29:52 -05:00
58e89b3024
Add floating point deconstruction helpers
2016-09-30 14:29:52 -05:00
f77ca915c4
add an opt-in cargo feature to build intrinsics from compiler-rt source
...
closes #63
cc #66
2016-09-29 16:06:24 -05:00
8e161a791a
Expand and refactor teting infrastructure
...
This commit moves over most of the testing infrastructure to in-tree docker
images that are all dispatched to from Travis (no other test configuration).
This allows versioning modifications to the test infrastructure as well as the
code itself. Additionally separate docker images allows for easy modification of
one without worrying about tampering of others as well as easy addition of new
targets by simply adding a new `Dockerfile`.
Additionally this commit bundles the master version of the `compiler-rt` source
repository from `llvm-mirror/compiler-rt` to test against. The compiler-rt
library itself is compiled as a `cdylib` which is then dynamically located at
runtime and we look for symbols in. There's a few hoops here, but they currently
get the job done.
All tests now execute against both gcc_s and compiler-rt, and this
testing strategy is now all hidden behind a macro as well (refactoring
all existing tests along the way).
2016-09-28 22:09:55 -07:00
c56a3f8a6e
Auto merge of #73 - japaric:weak, r=japaric
...
put weak mem* symbols behind an opt-in Cargo feature
closes #64
cc #66
2016-09-22 23:59:48 +00:00
b7fa7d3a59
put weak mem* symbols behind an opt-in Cargo feature
...
closes #64
cc #66
2016-09-22 18:02:25 -05:00
d5e38beef9
Auto merge of #71 - japaric:aeabi-aliases, r=japaric
...
add missing aeabi aliases
2016-09-22 22:55:43 +00:00
d34be6ab4e
add missing aeabi aliases
2016-09-22 17:02:36 -05:00
384c48ce9b
don't test always against gcc_s
...
instead test half of the time against gcc_s and the other half test
against the native operation (\*).
(\*) Not all the targets have available a native version of the
intrinsics under test. On those targets we'll end up testing our
implementation against itself half of the time. This is not much of a
problem because we do several quickcheck runs per intrinsic.
2016-09-21 21:38:06 -05:00
337bd7e209
armhf: don't compare our impls against gcc_s
2016-09-21 21:14:38 -05:00
225d4c9d45
avoid division overflow in idivmod test
2016-09-17 21:10:41 -05:00
9493c37455
test our implementations against gcc_s
...
if it exposes the same intrinsics that we implement -- gcc_s doesn't
implement all the intrinsics for all the architectures.
closes #65
2016-09-16 15:53:14 -05:00
2e561b3869
use utility function to compare the repr of floats
...
follow up of #43
2016-08-21 11:24:58 -05:00
c427112561
Merge pull request #49 from japaric/edge
...
qc: bias towards generation of edge cases
2016-08-21 11:03:44 -05:00
da53b70d34
Use mem::swap to swap variables
2016-08-20 22:30:02 -05:00
fc0cdeb7f1
qc: bias towards generation of edge cases
2016-08-20 22:17:50 -05:00
ef16de3d37
Implement soft float add builtins
2016-08-20 16:06:02 -05:00
6f0d50ebea
Move integer functions to separate module
2016-08-20 15:55:06 -05:00
8603e64554
Merge pull request #44 from Amanieu/sdiv
...
Add signed division functions
2016-08-19 09:16:53 -05:00
b9e916c769
Add tests for ARM division builtins
2016-08-19 12:13:02 +01:00
f919660be5
Add signed division functions
2016-08-19 12:13:02 +01:00
5e47ca2699
rename muldi4 to muldi3
...
closes #42
2016-08-18 21:27:16 -05:00
aab3a10a9d
Disable mangling for msvc stack builtins
...
This prevents linker errors in test builds due to the `jmp` instruction in __alloca
2016-08-17 12:55:15 -05:00
c9bff743cd
Use `jmp` to implement __alloca fallthrough
2016-08-17 12:11:38 -05:00
b6091ababa
Add x86_64 builtins
2016-08-16 19:55:17 -05:00
ac352a3512
add core_intrinsics feature gate
2016-08-15 21:08:04 -05:00
55eb1b71a0
remove unused feature gates
2016-08-14 22:08:36 -05:00
8965b061c2
use rlibc crate instead
2016-08-14 22:03:19 -05:00
fdbb7883da
exclude windows and macos
2016-08-14 22:01:15 -05:00
265939fc00
add weak `memcpy` et al symbols
...
closes #28
2016-08-14 22:00:31 -05:00
45aec943d3
use wrapping_add in muldi4
2016-08-13 18:47:39 -05:00
f3eb08a96d
fix a bug in udivmoddi4
2016-08-13 18:45:00 -05:00
69376af183
quickcheck: better generation of input arguments
...
closes #31
2016-08-13 16:58:44 -05:00
97217e3ad9
Fix inline assembly to work on ARMv6-M
2016-08-13 17:29:38 +01:00
acceaa95f1
Various changes
2016-08-13 09:56:40 +01:00
0a0a8f37c9
Remove the absv implementation, it isn't needed by Rust
2016-08-13 09:54:58 +01:00
eb206842f2
address second row of review
2016-08-11 01:45:48 -05:00
1c22269948
don't transmute in test
2016-08-11 01:42:34 -05:00
3ff25d956a
put div intrinsics in their own module + some docs
2016-08-11 01:26:27 -05:00
a84579d3c1
refactor: use Int traits, words -> U64, fmt
2016-08-11 01:15:51 -05:00
d6be95a3bb
note about calling convention
2016-08-11 00:36:19 -05:00
a2fc4da7a5
move return into if branches
2016-08-11 00:33:27 -05:00
0915e15c9a
*mut T -> Option<&mut T>
2016-08-11 00:31:49 -05:00
cc9d3e8e5f
port __udivmodsi4, aeabi_uidivmod and udivsi3
...
also rewrite these last two new aeabi intrinsics as naked functions
2016-08-11 00:26:16 -05:00
950564607d
port udivmoddi4 and __aeabi_uldivmod
2016-08-11 00:26:12 -05:00
est31
Tim Neumann
Oliver Middleton
Alex Crichton
Jorge Aparicio
Jorge Aparicio
bors
Wilfried Chauveau
James Duley
Vadzim Dambrouski
Matt Ickstadt
Christopher Serr
Matt Ickstadt
homunkulus
Amanieu d'Antras