[Bug target/70063] New: msp430 stack corruption for naked functions
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70063
Bug ID: 70063
Summary: msp430 stack corruption for naked functions
Product: gcc
Version: 5.3.0
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: target
Assignee: unassigned at gcc dot gnu.org
Reporter: awygle at gmail dot com
Target Milestone: ---
Created attachment 37853
--> https://gcc.gnu.org/bugzilla/attachment.cgi?id=37853&action=edit
A nearly minimal example of code producing the bug
When compiling msp430 code including functions which have both arguments and
the naked attribute using -O0, the stack is corrupted and the return pointer is
lost.
This looks like an msp430-style retread of
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=54220 to me.
Code:
# 1 "bug.c"
# 1 ""
# 1 ""
# 1 "bug.c"
__attribute__((naked))
void buggy(int* arg1, int* arg2) {
asm volatile("sub #4, r1");
(*arg1) = 2;
(*arg2) = 3;
asm volatile ("nop");
asm volatile ("nop");
asm volatile ("nop");
asm volatile ("nop");
asm volatile ("nop");
asm volatile ("ret");
}
int main() {
int x, y;
x = 0;
y = 1;
buggy(&x, &y);
while(1) {}
}
Compiling:
[awygle %] msp430-elf-gcc -v -Wall -Wextra -fno-strict-aliasing -fwrapv
-fno-aggressive-loop-optimizations -mmcu=msp430fr5969 bug.c -o bug.elf
-save-temps
Using built-in specs.
COLLECT_GCC=msp430-elf-gcc
COLLECT_LTO_WRAPPER=/home/awygle/toolchain/install/libexec/gcc/msp430-elf/5.3.0/lto-wrapper
Target: msp430-elf
Configured with: ../gcc-5.3.0/configure
--prefix=/home/awygle/toolchain/install/ --target=msp430-elf --without-headers
--with-newlib --enable-languages=c,c++ --disable-multilib -v
Thread model: single
gcc version 5.3.0 (GCC)
COLLECT_GCC_OPTIONS='-v' '-Wall' '-Wextra' '-fno-strict-aliasing' '-fwrapv'
'-fno-aggressive-loop-optimizations' '-mmcu=msp430fr5969' '-o' 'bug.elf'
'-save-temps'
/home/awygle/toolchain/install/libexec/gcc/msp430-elf/5.3.0/cc1 -E -quiet -v
bug.c -mmcu=msp430fr5969 -Wall -Wextra -fno-strict-aliasing -fwrapv
-fno-aggressive-loop-optimizations -fpch-preprocess -o bug.i
ignoring nonexistent directory
"/home/awygle/toolchain/install/lib/gcc/msp430-elf/5.3.0/../../../../msp430-elf/sys-include"
#include "..." search starts here:
#include <...> search starts here:
/home/awygle/toolchain/install/lib/gcc/msp430-elf/5.3.0/include
/home/awygle/toolchain/install/lib/gcc/msp430-elf/5.3.0/include-fixed
/home/awygle/toolchain/install/lib/gcc/msp430-elf/5.3.0/../../../../msp430-elf/include
End of search list.
COLLECT_GCC_OPTIONS='-v' '-Wall' '-Wextra' '-fno-strict-aliasing' '-fwrapv'
'-fno-aggressive-loop-optimizations' '-mmcu=msp430fr5969' '-o' 'bug.elf'
'-save-temps'
/home/awygle/toolchain/install/libexec/gcc/msp430-elf/5.3.0/cc1 -fpreprocessed
bug.i -quiet -dumpbase bug.c -mmcu=msp430fr5969 -auxbase bug -Wall -Wextra
-version -fno-strict-aliasing -fwrapv -fno-aggressive-loop-optimizations -o
bug.s
GNU C11 (GCC) version 5.3.0 (msp430-elf)
compiled by GNU C version 4.9.2, GMP version 6.0.0, MPFR version
3.1.2-p11, MPC version 1.0.3
GGC heuristics: --param ggc-min-expand=100 --param ggc-min-heapsize=131072
GNU C11 (GCC) version 5.3.0 (msp430-elf)
compiled by GNU C version 4.9.2, GMP version 6.0.0, MPFR version
3.1.2-p11, MPC version 1.0.3
GGC heuristics: --param ggc-min-expand=100 --param ggc-min-heapsize=131072
Compiler executable checksum: 4cbf41f67559bdb236a322a0045966e6
COLLECT_GCC_OPTIONS='-v' '-Wall' '-Wextra' '-fno-strict-aliasing' '-fwrapv'
'-fno-aggressive-loop-optimizations' '-mmcu=msp430fr5969' '-o' 'bug.elf'
'-save-temps'
/home/awygle/toolchain/install/lib/gcc/msp430-elf/5.3.0/../../../../msp430-elf/bin/as
--traditional-format -mP -mmcu=msp430fr5969 -md -o bug.o bug.s
COMPILER_PATH=/home/awygle/toolchain/install/libexec/gcc/msp430-elf/5.3.0/:/home/awygle/toolchain/install/libexec/gcc/msp430-elf/5.3.0/:/home/awygle/toolchain/install/libexec/gcc/msp430-elf/:/home/awygle/toolchain/install/lib/gcc/msp430-elf/5.3.0/:/home/awygle/toolchain/install/lib/gcc/msp430-elf/:/home/awygle/toolchain/install/lib/gcc/msp430-elf/5.3.0/../../../../msp430-elf/bin/
LIBRARY_PATH=/home/awygle/toolchain/install/lib/gcc/msp430-elf/5.3.0/:/home/awygle/toolchain/install/lib/gcc/msp430-elf/5.3.0/../../../../msp430-elf/lib/
COLLECT_GCC_OPTIONS='-v' '-Wall' '-Wextra' '-fno-strict-aliasing' '-fwrapv'
'-fno-aggressive-loop-optimizations' '-mmcu=msp430fr5969'
[Bug target/70063] msp430 stack corruption for naked functions
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70063 --- Comment #1 from Andrew Wygle --- A quick correction - as my example above accidentally shows, compiling with optimizations isn't a guarantee of avoiding this problem, it just sometimes helps if the arguments are unused, as the optimizer will remove the attempt to save them.
[Bug target/70713] New: msp430 interrupt attribute prevents overriding weak symbols
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70713
Bug ID: 70713
Summary: msp430 interrupt attribute prevents overriding weak
symbols
Product: gcc
Version: 5.3.0
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: target
Assignee: unassigned at gcc dot gnu.org
Reporter: awygle at gmail dot com
Target Milestone: ---
Because the interrupt attribute in the msp430 port creates a .word directive
inside of the generated assembly, thus:
test.c:
__attribute__((weak, interrupt(USCI_A1_VECTOR)))
void TEST_WEAKNESS(void) {
UCA1TXBUF = 'e';
}
msp430-elf-gcc -mmcu=msp430fr6989 -S test.c
test.s:
.file "test.c"
.text
.balign 2
.weak TEST_WEAKNESS
.section __interrupt_vector_43,"ax",@progbits
.word TEST_WEAKNESS
.text
TEST_WEAKNESS:
; start of function
; attributes: interrupt
; framesize_regs: 0
; framesize_locals: 0
; framesize_outgoing: 0
; framesize: 0
; elim ap -> fp 2
; elim fp -> sp 0
; saved regs:(none)
; start of prologue
; end of prologue
MOV.W #101, &UCA1TXBUF
NOP
; start of epilogue
RETI
.size TEST_WEAKNESS, .-TEST_WEAKNESS
When we compile and attempt to link with a non-weak symbol, thus:
test2.c:
__attribute__((interrupt(USCI_A1_VECTOR)))
void TEST_WEAKNESS(void) {
UCA1TXBUF = 'E';
}
msp430-elf-gcc -mmcu=msp430fr6989 -S test2.c
test2.s:
.file "test2.c"
.text
.balign 2
.global TEST_WEAKNESS
.section __interrupt_vector_43,"ax",@progbits
.word TEST_WEAKNESS
.text
TEST_WEAKNESS:
; start of function
; attributes: interrupt
; framesize_regs: 0
; framesize_locals: 0
; framesize_outgoing: 0
; framesize: 0
; elim ap -> fp 2
; elim fp -> sp 0
; saved regs:(none)
; start of prologue
; end of prologue
MOV.W #69, &UCA1TXBUF
NOP
; start of epilogue
RETI
.size TEST_WEAKNESS, .-TEST_WEAKNESS
The error:
msp430-elf-gcc -mmcu=msp430fr6989 test.s test2.s -o test.elf
msp430-elf/bin/ld: test.elf section `__interrupt_vector_43' will not fit in
region `VECT43'
msp430-elf/bin/ld: region `VECT43' overflowed by 2 bytes
collect2: error: ld returned 1 exit status
I was able to get this to work by making the section lines part of a comdat
group, so that
.section __interrupt_vector_43,"ax",@progbits
becomes
.section __interrupt_vector_43,"axG",@progbits,TEST_WEAKNESS,comdat
in both cases. There seem to be no ill-effects from doing this (provided the
group name is based on the function name), but I lack the gcc experience to say
for sure, or to cause such code to be emitted.
[Bug target/70063] msp430 stack corruption for naked functions
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70063 --- Comment #3 from Andrew Wygle --- This appears to be fixed in GCC 6.1
[Bug target/80993] New: [msp430] __attribute__((interrupt)) should imply __attribute__((used))
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=80993 Bug ID: 80993 Summary: [msp430] __attribute__((interrupt)) should imply __attribute__((used)) Product: gcc Version: 6.2.0 Status: UNCONFIRMED Severity: normal Priority: P3 Component: target Assignee: unassigned at gcc dot gnu.org Reporter: awygle at gmail dot com Target Milestone: --- Enabling link-time optimization on the MSP430 causes any interrupt handlers which are not called directly to be cheerfully eaten. Thus any interrupt causes the processor to execute essentially random data, resulting in uncontrolled jumps and general badness. This problem can be resolved with __attribute__((used)), which should be implied by __attribute__((interrupt)).
[Bug target/78554] New: Internal Compiler Error in msp430 target with -mlarge, -O{s123}
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=78554
Bug ID: 78554
Summary: Internal Compiler Error in msp430 target with -mlarge,
-O{s123}
Product: gcc
Version: 6.2.0
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: target
Assignee: unassigned at gcc dot gnu.org
Reporter: awygle at gmail dot com
Target Milestone: ---
When compiling code for an MSP430 using large model (-mlarge) and any
optimization above 0, assigning the address of a global variable to an __int20
unsigned (__UINTPTR_TYPE__) which is part of a structure causes an internal
compiler error, as in the following .i file:
# 1 "test.c"
# 1 ""
# 1 ""
# 1 "test.c"
unsigned char test_val = 0;
typedef __int20 unsigned reg_t;
struct holds_reg {
reg_t r0;
};
struct holds_reg ex = { 0 };
int main() {
ex.r0 = (reg_t)(&test_val);
return 0;
}
The invocation of msp430-elf-gcc:
msp430-elf-gcc -mmcu=msp430fr5969 -mlarge -O1 -save-temps test.c
test.c: In function ‘main’:
test.c:14:1: error: unrecognizable insn:
}
^
(insn 6 5 8 2 (set (mem/c:HI (reg/f:PSI 25) [2 ex.r0+0 S2 A16])
(subreg:HI (symbol_ref:PSI ("test_val") [flags 0x2] ) 0)) test.c:12 -1
(nil))
test.c:14:1: internal compiler error: in extract_insn, at recog.c:2287
0x8aa29a _fatal_insn(char const*, rtx_def const*, char const*, int, char
const*)
../../gcc-6.2.0/gcc/rtl-error.c:108
0x8aa2c9 _fatal_insn_not_found(rtx_def const*, char const*, int, char const*)
../../gcc-6.2.0/gcc/rtl-error.c:116
0x882087 extract_insn(rtx_insn*)
../../gcc-6.2.0/gcc/recog.c:2287
0x6e1ba5 instantiate_virtual_regs_in_insn
../../gcc-6.2.0/gcc/function.c:1582
0x6e1ba5 instantiate_virtual_regs
../../gcc-6.2.0/gcc/function.c:1950
0x6e1ba5 execute
../../gcc-6.2.0/gcc/function.c:1999
Please submit a full bug report,
with preprocessed source if appropriate.
Please include the complete backtrace with any bug report.
See <http://gcc.gnu.org/bugs.html> for instructions.
Some things which don't cause this error include assigning the address of a
variable which is local to main or assigning the address directly to a __int20
unsigned which is not part of a struct. The error does not occur on -msmall or
on -O0.
[Bug target/78818] New: msp430 persistent attribute is not applied correctly in some cases
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=78818
Bug ID: 78818
Summary: msp430 persistent attribute is not applied correctly
in some cases
Product: gcc
Version: 6.2.0
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: target
Assignee: unassigned at gcc dot gnu.org
Reporter: awygle at gmail dot com
Target Milestone: ---
The MS430 variable attribute "persistent" places variables in a special section
.persistent so that they will not be re-initialized on reset and so that the
linker can place them in non-volatile storage. Variables with this attribute
aren't placed in this section when:
1) the variables are specified as "static" AND are not initialized or are
initialized to 0. In this case the variables are placed in the .bss section.
2) the program is compiled with the -fdata-sections optimization option. In
this case the variables are placed in either .bss.foo or .text.foo.
An example program and two compilations:
test.i:
# 1 "test.c"
# 1 ""
# 1 ""
# 1 "test.c"
__attribute__((persistent)) int persist = 1;
__attribute__((persistent)) int global_persist = 0;
static __attribute__((persistent)) int static_persist = 1;
static __attribute__((persistent)) int static_nopersist = 0;
void main() {
while(1);
}
Compilation without -fdata-sections:
% msp430-elf-gcc -msim -c test.i -o test.o
% msp430-elf-objdump -x --syms test.o
test.o: file format elf32-msp430
test.o
architecture: MSP430X, flags 0x0011:
HAS_RELOC, HAS_SYMS
start address 0x
Sections:
Idx Name Size VMA LMA File off Algn
0 .text 0004 0034 2**1
CONTENTS, ALLOC, LOAD, RELOC, READONLY, CODE
1 .data 0038 2**0
CONTENTS, ALLOC, LOAD, DATA
2 .bss 0002 0038 2**1
ALLOC
3 .persistent 0006 0038 2**1
CONTENTS, ALLOC, LOAD, DATA
4 .comment 0012 003e 2**0
CONTENTS, READONLY
5 .MSP430.attributes 0017 0050 2**0
CONTENTS, READONLY
SYMBOL TABLE:
ldf *ABS* test.c
ld .text .text
ld .data .data
ld .bss .bss
ld .persistent .persistent
0004 l O .persistent0002 static_persist
l O .bss 0002 static_nopersist
ld .comment .comment
ld .MSP430.attributes .MSP430.attributes
g O .persistent0002 persist
0002 g O .persistent0002 global_persist
*UND* __crt0_init_bss
g F .text 0004 main
RELOCATION RECORDS FOR [.text]:
OFFSET TYPE VALUE
0002 R_MSP430X_ABS16 .L2
Compilation with -fdata-sections:
% msp430-elf-gcc -msim -fdata-sections -c test.i -o test.o
% msp430-elf-objdump -x --syms test.o
test.o: file format elf32-msp430
test.o
architecture: MSP430X, flags 0x0011:
HAS_RELOC, HAS_SYMS
start address 0x
Sections:
Idx Name Size VMA LMA File off Algn
0 .text 0004 0034 2**1
CONTENTS, ALLOC, LOAD, RELOC, READONLY, CODE
1 .data 0038 2**0
CONTENTS, ALLOC, LOAD, DATA
2 .bss 0038 2**0
ALLOC
3 .data.persist 0002 0038 2**1
CONTENTS, ALLOC, LOAD, DATA
4 .bss.global_persist 0002 003a 2**1
ALLOC
5 .data.static_persist 0002 003a 2**1
CONTENTS, ALLOC, LOAD, DATA
6 .bss.static_nopersist 0002 003c 2**1
ALLOC
7 .comment 0012 003c 2**0
CONTENTS, READONLY
8 .MSP430.attributes 0017 004e 2**0
CONTENTS, READONLY
SYMBOL TABLE:
ldf *ABS* test.c
ld .text .text
ld .data .data
ld .bss .bss
ld .data.persist .data.persist
ld .bss.global_persist .bss.global_persist
ld .data.static_persist .data.static_persist
l O .data.static_persist 0002 static_persist
ld .bss.static_nopersist .bss.static_nopersist
l O .bss.static_nopersist 0002 static_nopersist
ld .co
[Bug target/78838] New: msp430 option -mcode-region=either, -ffunction-sections, and interrupt function attributes cause incorrect section to be created
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=78838
Bug ID: 78838
Summary: msp430 option -mcode-region=either,
-ffunction-sections, and interrupt function attributes
cause incorrect section to be created
Product: gcc
Version: 6.2.0
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: target
Assignee: unassigned at gcc dot gnu.org
Reporter: awygle at gmail dot com
Target Milestone: ---
On the msp430 target, when using -mcode-region with -ffunction-sections (as is
semi-suggested by the documentation), any functions which are marked with the
interrupt attribute and which also specify an interrupt number are placed into
an incorrect section ".either.lowtext", which is clearly nonsensical. An
example is below.
interrupt_bug.i:
# 1 "interrupt_bug.c"
# 1 ""
# 1 ""
# 1 "interrupt_bug.c"
__attribute__((interrupt(2))) void causes_error(void) {
}
void main() {
while(1);
}
Compiling with -ffunction-sections:
% msp430-elf-gcc -msim -mlarge -mcode-region=either -ffunction-sections
-save-temps -c interrupt_bug.c -o interrupt_bug.o
% msp430-elf-gcc -msim -mlarge -mcode-region=either -ffunction-sections
-save-temps interrupt_bug.o -o interrupt_bug.elf
/home/awygle/toolchain/install/bin/../lib/gcc/msp430-elf/6.2.0/../../../../msp430-elf/bin/ld:
error: no section named .lower.lowtext or .upper.lowtext in linker script
% msp430-elf-objdump --syms interrupt_bug.o
interrupt_bug.o: file format elf32-msp430
SYMBOL TABLE:
ldf *ABS* interrupt_bug.c
ld .text .text
ld .data .data
ld .bss .bss
ld .either.lowtext .either.lowtext
ld __interrupt_vector_2 __interrupt_vector_2
ld .either.text.main .either.text.main
ld .comment .comment
ld .MSP430.attributes .MSP430.attributes
g F .either.lowtext0004 causes_error
g F .either.text.main 0004 main
Note that main() is in the section .either.text.main, while causes_error() is
in the section .either.lowtext.
Compiling without -ffunction-sections:
% msp430-elf-gcc -msim -mlarge -mcode-region=either -save-temps -c
interrupt_bug.c -o interrupt_bug.o
% msp430-elf-gcc -msim -mlarge -mcode-region=either -save-temps interrupt_bug.o
-o interrupt_bug.elf
% msp430-elf-objdump --syms interrupt_bug.o
interrupt_bug.o: file format elf32-msp430
SYMBOL TABLE:
ldf *ABS* interrupt_bug.c
ld .text .text
ld .data .data
ld .bss .bss
ld .lowtext .lowtext
ld __interrupt_vector_2 __interrupt_vector_2
ld .either.text .either.text
ld .comment .comment
ld .MSP430.attributes .MSP430.attributes
g F .lowtext 0004 causes_error
g F .either.text 0004 main
Now causes_error() is in the .lowtext section, as expected.
[Bug target/78849] New: ICE on initialization of global struct containing __int20 array at varasm.c:4968
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=78849
Bug ID: 78849
Summary: ICE on initialization of global struct containing
__int20 array at varasm.c:4968
Product: gcc
Version: 6.2.0
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: target
Assignee: unassigned at gcc dot gnu.org
Reporter: awygle at gmail dot com
Target Milestone: ---
The following code (constructor_bug.i) produces an internal compiler error when
compiled for the msp430-elf target:
# 1 "constructor_bug.c"
# 1 ""
# 1 ""
# 1 "constructor_bug.c"
struct causes_problems {
__int20 array[2];
char value;
};
struct causes_problems instance = {
{
0,
1,
},
0,
};
void main(void) {
while(1);
}
% msp430-elf-gcc -msim constructor_bug.i
constructor_bug.c:17:1: internal compiler error: in
output_constructor_regular_field, at varasm.c:4968
}
^
0xb06039 output_constructor_regular_field
../../gcc-6.2.0/gcc/varasm.c:4968
0xb06039 output_constructor
../../gcc-6.2.0/gcc/varasm.c:5276
0xb05254 assemble_variable_contents
../../gcc-6.2.0/gcc/varasm.c:2062
0xb0bdca assemble_variable(tree_node*, int, int, int)
../../gcc-6.2.0/gcc/varasm.c:2238
0xb0fadd varpool_node::assemble_decl()
../../gcc-6.2.0/gcc/varpool.c:582
0x600403 output_in_order
../../gcc-6.2.0/gcc/cgraphunit.c:2231
0x6006a3 symbol_table::compile()
../../gcc-6.2.0/gcc/cgraphunit.c:2468
0x6025d9 symbol_table::finalize_compilation_unit()
../../gcc-6.2.0/gcc/cgraphunit.c:2564
Please submit a full bug report,
with preprocessed source if appropriate.
Please include the complete backtrace with any bug report.
See <http://gcc.gnu.org/bugs.html> for instructions.
% msp430-elf-gcc -v
Using built-in specs.
COLLECT_GCC=msp430-elf-gcc
COLLECT_LTO_WRAPPER=/home/awygle/toolchain/install/libexec/gcc/msp430-elf/6.2.0/lto-wrapper
Target: msp430-elf
Configured with: ../gcc-6.2.0/configure
--prefix=/home/awygle/toolchain/install/ --target=msp430-elf --without-headers
--with-newlib --enable-languages=c,c++ -v
Thread model: single
gcc version 6.2.0 (GCC)
As far as I can tell this is a minimal test case, by which I mean that any of
the following prevents the error:
* Reducing the length of the array to 1
* Removing the "value" field from the struct
* Moving the array field to the end of the struct
* Not initializing the struct
* Moving the struct variable definition and initialization into the main()
function
I filed this under "target" but I suspect it would apply to any
non-byte-aligned __intN type - however, only one of those seems to exist right
now and it's in the msp430 target. I tried to reproduce this bug on the AVR
target using __int24 without success, but my avr-gcc version is 4.8.2 rather
than 6.2.
Thanks
[Bug target/78818] msp430 persistent attribute is not applied correctly in some cases
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=78818 --- Comment #1 from Andrew Wygle --- Compiling with -mdata-region=either also causes this problem, with variables ending up in either .either.data or .either.bss.
