67786705a5
Update M2libc for more efficient RISC-V syscalls
2021-10-21 21:10:53 -04:00
ffe4e5f66e
better validate arguments passed
2021-10-04 20:49:44 -04:00
ae9e700592
Release_1.9.0
2021-10-03 09:34:22 -04:00
Andrius Štikonas
56ee526c4c
Implement support for large immediates on RISC-V.
2021-10-03 11:37:03 +01:00
5e74995c13
Catch truncated function calls and truncated array statements
2021-10-02 22:02:39 -04:00
0577243444
Fix Problem with global arrays #18
2021-10-02 10:38:04 -04:00
4af9b79c1b
restore removed test checksums
2021-10-02 08:24:07 -04:00
Andrius Štikonas
bbf67601e1
Add tests for riscv64.
2021-10-02 13:13:15 +01:00
Andrius Štikonas
ae48dbd6cb
Add riscv64 port.
2021-10-02 11:59:21 +01:00
7fb9d682ec
Update M2libc to enable RISC-V Development
2021-09-30 20:10:28 -04:00
745998ac13
Fix global array initialization to allocate the correct amount of memory
2021-06-26 11:58:13 -04:00
99df87bc0d
Breakup program to reduce complexity and risk of errors
2021-06-26 11:51:54 -04:00
b37dfc7b12
Report proper error message for negative global array values. Thank you melg8
2021-06-26 10:53:45 -04:00
51dfec1791
Fixed Global buffer overflow in collect_weird_string thanks to Melg8
2021-06-23 18:57:56 -04:00
a3c14bcd1b
Fixed Memory leak, thanks melg8
2021-06-23 07:58:11 -04:00
8bc09f2b2e
Fixed hang on large global arrays. Thanks melg8
2021-06-23 07:32:40 -04:00
da542698e7
Fix all special cases where line number was wrong
2021-05-28 23:27:17 -04:00
cece07145c
Update checksums to reflect new cc_types.c code in test1000
2021-05-02 10:52:29 -04:00
d80d8a652b
Move more functionality into M2libc
2021-04-03 22:41:52 -04:00
46cf81af83
Replace file_print with fputs
2021-04-03 18:56:55 -04:00
b0b440c2df
Update M2libc to fix fflushing of buffers at exit or return from main
2021-02-21 21:00:34 -05:00
163dab268c
Merge remote-tracking branch 'deesix/TMP'
2021-02-17 23:01:23 -05:00
37ca8e3df9
Remove test/common_x86 files and switch x86 to M2libc
2021-02-17 22:01:28 -05:00
32ef09f964
Update M2libc and update known_issues.org
2021-02-06 13:40:28 -05:00
fc4c1998aa
Remove test/common_knight/ files and switch knight-posix to M2libc
2021-02-06 00:45:32 -05:00
a9d88e8e6c
Add support for -D variable and -D variable=value
2021-02-05 21:34:00 -05:00
8007ae9ce6
Add support for #if defined(__M2__) to enable M2-Planet specific functionality
2021-02-05 20:39:54 -05:00
62eedce858
Remove test/common_aarch64 files and switch AArch64 to M2libc
2021-02-05 15:54:38 -05:00
a65619dbf4
Remove test/common_knight/*native files and switch knight-native to M2libc
2021-02-03 19:47:18 -05:00
ba1fc44f29
Remove test/common_armv7l files and switch armv7l to M2libc
2021-02-03 07:30:15 -05:00
695672985e
Remove test/common_amd64 files and switch amd64 to M2libc
2021-02-03 00:32:47 -05:00
6f8e2eb905
Removed need for fixup
2021-02-02 23:39:27 -05:00
f865e47646
@@@ tests: aarch64 test0100 using debug ELF header
...
It was using the non-debug one but the rest of the compilation chain is
done with debug in mind.
2021-01-31 00:00:00 +01:00
1a9bbe6e63
@@@ tests: same --debug behaviour for all scripts of the same test
...
Until now, a few scripts compiled the test program with/without debug,
differently to the rest.
2021-01-31 00:00:00 +01:00
c5068b227a
Add support for &global_variable
2021-01-28 22:22:08 -05:00
f2094a9e5b
Migrate knight-posix to M2libc
2021-01-22 00:35:43 -05:00
3c6b23f734
Update M2libc to remove duplicate opens
2021-01-21 22:41:23 -05:00
818784228c
Use M2libc in all x86 tests < test1000
2021-01-21 00:00:00 +01:00
e660caca73
Use M2libc in all armv7l tests < test1000
2021-01-21 00:00:00 +01:00
7860378b75
Use M2libc in all amd64 tests < test1000
2021-01-21 00:00:00 +01:00
55db771e29
tests: get proof from test0025 stderr (and other minor fixes)
...
We were checking an empty proof because the test writes to stderr and we
were redirecting stdout to the file. With the correct redirection the hash
is different, so proof.answer is also adjusted.
The x86 test was calling blood-elf with the 64bits flag. Removed to ask for
32bits behaviour.
For consistency, --BaseAddress is now using the usual value for the arch.
Also, unused input file is removed here.
2021-01-17 00:00:00 +01:00
bcd0f49859
Use M2libc in all AArch64 tests < test1000
...
Finally test0021, now that M2libc has getcwd.
See b83086642e
2021-01-15 00:00:00 +01:00
4def464475
Merge remote-tracking branch 'yt/array'
2021-01-10 20:07:38 -05:00
768f229c1d
Updated M2libc
2021-01-10 18:29:28 -05:00
0e3d5b609a
Add support for global char foo[12]; definitions
...
- only enabled in --bootstrap-mode
- allocates space for both the char array and the pointer to the start of
the array
2021-01-10 23:01:08 +00:00
b83086642e
Use M2libc in nearly all AArch64 tests
...
- disables --bootstrap-mode where it was enabled, except for test1000
which needs to build in bootstrap-mode to check that M2-Planet can be
compiled by cc_*
- the only hold-out is test0021 which needs getcwd which is missing in
M2libc
- note that test0106 needs a fix to fgetc or it will never notice it has
reached the end of stdin
2021-01-10 22:02:29 +00:00
dd1f5db9f1
Port AArch64 test0008 to M2libc
2021-01-10 17:50:04 +00:00
c4d329b6cc
Fixup libc-core -> libc-full
...
Causes a segfault due to uninitialised pointer in the FILE struct
2021-01-10 17:00:12 +00:00
45685d0c30
Update first few tests to use M2libc
2021-01-10 16:47:13 +00:00
09acd6253d
Implement simple macro expansion
...
- no longer generate CONSTANT lines for #defined constants
- "#define FOO 1" turns "return FOO;" into "return 1;"
- Macros expanding into a macro is not supported:
#define FOO 1
#define BAR FOO
will expand BAR to FOO, not to 1
- Expansion inside #if conditions is not supported.
- Function-like macros are not supported.
2021-01-09 17:39:48 +00:00
a8551f2fcd
Implement --bootstrap-mode for preprocessor
...
- // and # lines always appear in the result of read_all_tokens
- in bootstrap mode, // tokens (but not the subsequent lines) and #
lines are stripped. no preprocessing happens.
- in non-bootstrap mode, // lines are stripped and # lines stay for the
preprocessing phase
- updates tests to pass --bootstrap-mode when necessary
2021-01-09 17:38:50 +00:00
6f2cebc4ca
Add a test for supported preprocessor features
2021-01-08 21:32:03 +00:00
4550eec19e
Extend macro #define support
...
- now supports #if defined(FOO)
- translation of #define FOO 1 to CONSTANT FOO 1 is still supported
2021-01-08 21:31:43 +00:00
3c4c09b95e
Added hacky version of #define support
2021-01-06 19:24:16 -05:00
25fd5e66c9
Another set of segfaults cleared out
2021-01-05 23:22:26 -05:00
bdd4e65580
Catch some basic segfaults
2021-01-05 23:00:58 -05:00
63bb59404d
Improve error reporting for unexpected non-newline
2021-01-05 22:56:30 +00:00
49fbe8942f
Add error checking for unexpected #endif
2021-01-05 22:56:30 +00:00
89c6513cd8
Add support for conditional inclusion
...
- support for #if, #elif, #else, #endif, but no macro expansion just
yet. Any macro is treated as undefined and expands to 0.
for example:
./bin/M2-Planet -E test.c
#if 1 == FOO
foo
#else
bar
#endif
results in
bar
2021-01-05 22:54:30 +00:00
cd96f65a23
Add macro directives to token_list
...
- change the lexer to not remove preprocessor directives
- preprocess() removes any lines starting with a macro directive that it
doesn't understand (which is currently all of them)
2021-01-05 22:53:39 +00:00
7b8a99db3c
Add a preprocessor-only mode
...
- Very useful for testing and debugging
- Add preprocess() stub and skip compilation in preprocessor mode
2021-01-05 22:53:04 +00:00
a18e0c1782
Add newline tokens to parse stream
...
- needed to determine start and end of macro directives
- the main parser doesn't need them, so strip out the newline tokens
before parsing to avoid changing it
2021-01-05 22:23:15 +00:00
921cc86ce6
First generation implemention of typedef
2021-01-02 22:00:02 -05:00
358b6cfb96
Fix modulus behavior that is wrong.
...
Thank you janneke
2021-01-01 16:39:27 -05:00
e451abd667
Fix e25c1995aa SCM MES regression
2021-01-01 15:55:34 -05:00
cc1a17a040
Revert "file.c: introduce output buffering for x86"
...
This reverts commit 5afeb0743e
2021-01-01 15:28:30 -05:00
07baf02efe
Looks like I got x86/AMD64 signed and unsigned division backwards.
...
Thanks for finding that janneke
2021-01-01 10:52:23 -05:00
c220d14928
Catching some segfaults
2020-12-31 09:53:05 -05:00
e25c1995aa
Make types more proper
2020-12-27 21:51:46 -05:00
0ed81de9e1
Merge remote-tracking branch 'mihi/io-buffering-x86'
2020-12-22 22:19:44 -05:00
4de11b1bc6
Update checksums to reflect new fclose
2020-12-21 22:26:22 -05:00
5afeb0743e
file.c: introduce output buffering for x86
...
Note that this is a backwards incompatible change: Unlike POSIX, exit()
will not flush all open streams, so you are responsible for flushing or
closing all your streams before successful exit.
Also, file.c will now also require calloc.c (and malloc.c).
Updated the test cases accordingly.
Only implemented and tested for x86.
2020-12-21 16:57:17 +01:00
91bab2ef98
Close files that were fopen:ed
2020-12-21 16:57:17 +01:00
6fe6f44a29
Make type information localized to the inside of statements
2020-12-19 16:35:33 -05:00
e7a0ecd244
Cleared out non-essential string primitives and added support for --max-string
2020-12-19 09:46:22 -05:00
d9504e3872
Allow M2-Planet to support larger outputs without hitting a segfault
2020-12-19 06:50:11 -05:00
968fdfea65
AArch64 support for unsigned comparison
2020-12-17 00:01:17 +00:00
e5befc4fee
armv7l support for unsigned comparisons
2020-12-15 21:09:26 -05:00
f2cae3d501
AMD64 support for unsigned comparisons
2020-12-15 20:28:04 -05:00
eaf3861954
Fix failing tests on x86
2020-12-15 19:02:32 -05:00
132e685f97
Fix x86 behavior on unsigned compares
2020-12-15 17:46:45 -05:00
9da48c5ece
match gcc behavior in regards to if(-1 > 0)
2020-12-15 07:24:07 -05:00
5c22aac02d
Change knight-posix HEAP behavior and leverage better relational choices for knight-posix
2020-12-15 06:02:38 -05:00
a63b8837c0
[AArch64] Use arithmetic right shift for signed
...
- as discussed with deesix, this is the proper fix
- the test1000 checksum changes as it uses the M2-Planet source as input
2020-12-05 14:59:16 +00:00
58b563435b
[AArch64] Use arithmetic right shift
...
- fixes a test failure if hex2 was compiled with M2-Planet
2020-12-04 19:37:29 +00:00
624a5ac92d
[AArch64] Fix LSEEK syscall number
2020-12-04 15:45:06 +00:00
0615223052
Broke tests into processor specific sets and updated the M1 test
2020-12-01 22:24:36 -05:00
67dbbbdcfc
Enable support for C multi-strings
2020-11-15 18:58:25 -05:00
57264eb157
Catch potential segfaulting behavior in generated binaries
2020-11-08 21:13:12 -05:00
39915de931
harmonize integer behavior when reading strings
2020-10-17 13:50:25 -04:00
c4a636c99d
Fix shift behavior to better match C standard
2020-10-16 22:07:06 -04:00
52114585bf
fix malloc to accept FreeBSD's behavior and fix failing binaries
2020-10-16 17:26:39 -04:00
028bdaf951
Update ELF headers and sha256 function to better support FreeBSD
2020-10-16 12:00:42 -04:00
87672ddfdb
Release 1.6.0
2020-04-30 22:35:22 -04:00
205c383c5f
lib: fflush stub
...
Paraphrasing OriansJ's IRC message while working on kaem:
GCC needs fflush(stdout) to get matching behavior, as M2-Planet doesn't
buffer. I guess we will need to make a fflush function (it'll do nothing
but return 0).
2020-03-01 18:55:32 +01:00
09783ac089
Fixed continue behavior and added a test to verify correct behavior
2020-02-28 18:48:58 -05:00
6343e48600
Major relocation of tests to better structure for future additions
2020-02-28 18:27:12 -05:00
b2ece54864
aarch64: do not generate a wasteful stack anymore
...
Until now, each element on the compiled program stack was 128 bits long,
half of them unused because only one 64 bits value was stored.
Now it's 64 bits long, so we don't waste all that memory.
We workaround the architectural alignment restriction of the SP register by
using a free regular register. X18 is for platform use so it seems a good
candidate for this task.
At _start we copy the value of SP into X18. SP is not used anymore. When a
definition refers to SP it doesn't mean it literaly now, because here we
redefine (without renaming) the involved M1 macro definitions to operate
on X18 (easier transition; abstraction). INIT_SP is introduced.
The function arguments are passed via stack, so the offset from the Base
Pointer ("depth") to each of them is different now. Changes both to
compiler code (cc_core.c) and libc reflect that the arguments are 8 bytes
(instead of 16 bytes) apart now. Note that SUB_X0_32 and SUB_X0_48 are
removed, because we only need 8, 16 and 24 bytes subtraction to reach the
arguments from asm libc functions. SUB_X0_8 and SUB_X0_24 are introduced.
2020-02-08 00:16:54 +01:00
b2b1e13b03
aarch64: test27 links to the usual 0x400000 base address
2020-02-05 13:20:58 +01:00
aad1198faf
Revised tests to include dwarf stubs
2020-02-01 17:50:26 -05:00
Jeremiah Orians
Andrius Štikonas
deesix
Sanne Wouda
Michael Schierl
nimaje