[Bug c/56473] New: Cross Compile Ubuntu x86_64 to Raspberry Pi Code Seg Faults When Executed
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56473
Bug #: 56473
Summary: Cross Compile Ubuntu x86_64 to Raspberry Pi Code Seg
Faults When Executed
Classification: Unclassified
Product: gcc
Version: 4.6.3
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: c
AssignedTo: unassig...@gcc.gnu.org
ReportedBy: tomd...@speakeasy.org
Created attachment 29548
--> http://gcc.gnu.org/bugzilla/attachment.cgi?id=29548
Script of the compile/link
I compiled and linked on Ubuntu x86_64 for the Raspberry Pi.
The compiler/linker completed with no errors.
int main() { return 0;}
When hello is executed on Raspberry Pi, I got a Segmentation fault.
If I compile on Ubuntu with -nostdlib and explicitly list the Raspberry Pi
crtxxx.o files and the Raspberry Pi libs, the resultant executable completes
without error on Raspberry Pi.
I attached a script of the compile actions on Ubuntu.
I believe this to be a problem with the crt.o files and the libs
distributed with arm-linux-gnueabihf-gcc-4.6 package.
The same package on Raspberry Pi works Ok.
On Ubuntu
> uname -a
Linux P9X79 3.2.0-38-generic #60-Ubuntu SMP Wed Feb 13 13:22:43 UTC 2013 x86_64
x86_64 x86_64 GNU/Linux
On Raspberry Pi
> uname -a
Linux raspberrypi 3.2.27+ #250 PREEMPT Thu Oct 18 19:03:02 BST 2012 armv6l
GNU/Linux
[Bug target/56473] Cross Compile Ubuntu x86_64 to Raspberry Pi Code Seg Faults When Executed
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56473 --- Comment #3 from Thomas Dean 2013-02-28 00:24:44 UTC --- On Ubuntu, I recompiled with -g On RaspberryPi, I ran gdb on the resulting executable tomdean@raspberrypi:~/p9x79/bug$ gdb hello GNU gdb (GDB) 7.4.1-debian Reading symbols from /home/tomdean/p9x79/bug/hello...done. (gdb) b main Breakpoint 1 at 0x8364: file hello.c, line 3. (gdb) r Starting program: /home/tomdean/p9x79/bug/hello Program received signal SIGSEGV, Segmentation fault. 0x000573a6 in ?? () (gdb) bt #0 0x000573a6 in ?? () #1 0x82ee in _start () I think the ubuntu crtxxx.o files and the unbunt arm-linux-gnueabihf-gcc-4.6 libs were not built correctly. diff says they are not the same as the RaspberryPi files. If I compile on ubuntu and link on RaspberryPi, all is well. If I compile on ubuntu and link, using the crtxxx.o and libs from RaspberryPi, all is well.
[Bug target/56473] Cross Compile Ubuntu x86_64 to Raspberry Pi Code Seg Faults When Executed
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56473
--- Comment #4 from Thomas Dean 2013-02-28
00:38:56 UTC ---
Looking again at gdb, _start from ubuntu is not the same as _start on
RaspberryPi, although both compiler versions are the same.
###
## build on ubuntu
## gdb on raspi
tomdean@raspberrypi:~/p9x79/bug$ gdb hello
GNU gdb (GDB) 7.4.1-debian
Reading symbols from /home/tomdean/p9x79/bug/hello...done.
(gdb) b main
Breakpoint 1 at 0x8364: file hello.c, line 3.
(gdb) r
Starting program: /home/tomdean/p9x79/bug/hello
Program received signal SIGSEGV, Segmentation fault.
0x000573a6 in ?? ()
(gdb) bt
#0 0x000573a6 in ?? ()
#1 0x82ee in _start ()
(gdb) disassemble/r _start
Dump of assembler code for function _start:
0x82e4 <+0>: 4f f0 00 0b mov.w r11, #0
0x82e8 <+4>: 4f f0 00 0e mov.w lr, #0
0x82ec <+8>: 5d f8 04 1b ldr.w r1, [sp], #4
0x82f0 <+12>:6a 46 mov r2, sp
0x82f2 <+14>:4d f8 04 2d str.w r2, [sp, #-4]!
0x82f6 <+18>:4d f8 04 0d str.w r0, [sp, #-4]!
0x82fa <+22>:df f8 14 c0 ldr.w r12, [pc, #20] ; 0x8310
<_start+44>
0x82fe <+26>:4d f8 04 cd str.w r12, [sp, #-4]!
0x8302 <+30>:04 48 ldr r0, [pc, #16] ; (0x8314 <_start+48>)
0x8304 <+32>:04 4b ldr r3, [pc, #16] ; (0x8318 <_start+52>)
0x8306 <+34>:ff f7 da ef blx 0x82bc <__libc_start_main>
0x830a <+38>:ff f7 e6 ef blx 0x82d8
0x830e <+42>:00 00 bd 83 ;
instruction: 0x83bd
0x8312 <+46>:00 00 5c 83 cmphi r12, #0
0x8316 <+50>:00 00 79 83 cmnhi r9, #0
0x831a <+54>:00 00 03 4b blmi0xc8322
End of assembler dump.
###
## compile on ubuntu
## link on raspi
## gdb on raspi
ubuntu> arm-linux-gnueabihf-gcc-4.6 -g -mcpu=arm1176jzf-s -mfpu=vfp
-mfloat-abi=hard -marm hello.c -Wl,-t -c -o hello.o
tomdean@raspberrypi:~/p9x79/bug$ arm-linux-gnueabihf-gcc-4.6 -g
-mcpu=arm1176jzf-s -mfpu=vfp -mfloat-abi=hard -marm hello.o -Wl,-t -o
~/work/hello
tomdean@raspberrypi:~/p9x79/bug$ gdb ~/work/hello
GNU gdb (GDB) 7.4.1-debian
Reading symbols from /home/tomdean/work/hello...done.
(gdb) b main
Breakpoint 1 at 0x8398: file hello.c, line 3.
(gdb) r
Starting program: /home/tomdean/work/hello
Breakpoint 1, main () at hello.c:3
3 return 0;
(gdb) disassemble/r _start
Dump of assembler code for function _start:
0x82e4 <+0>: 00 b0 a0 e3 mov r11, #0
0x82e8 <+4>: 00 e0 a0 e3 mov lr, #0
0x82ec <+8>: 04 10 9d e4 pop {r1}; (ldr r1,
[sp], #4)
0x82f0 <+12>:0d 20 a0 e1 mov r2, sp
0x82f4 <+16>:04 20 2d e5 push{r2}; (str r2, [sp,
#-4]!)
0x82f8 <+20>:04 00 2d e5 push{r0}; (str r0, [sp,
#-4]!)
0x82fc <+24>:10 c0 9f e5 ldr r12, [pc, #16] ; 0x8314
<_start+48>
0x8300 <+28>:04 c0 2d e5 push{r12} ; (str r12,
[sp, #-4]!)
0x8304 <+32>:0c 00 9f e5 ldr r0, [pc, #12] ; 0x8318
<_start+52>
0x8308 <+36>:0c 30 9f e5 ldr r3, [pc, #12] ; 0x831c
<_start+56>
0x830c <+40>:eb ff ff eb bl 0x82c0 <__libc_start_main>
0x8310 <+44>:f0 ff ff eb bl 0x82d8
0x8314 <+48>:0c 84 00 00 andeq r8, r0, r12, lsl #8
0x8318 <+52>:90 83 00 00 muleq r0, r0, r3
0x831c <+56>:ac 83 00 00 andeq r8, r0, r12, lsr #7
End of assembler dump.
[Bug libstdc++/43802] New: Tanh Returns Incorrect Value
In c and c++, tanh returns a different incorrect value.
test-tanh.c: gcc test-tanh.c -o test-tanh-c-lm
test-tanh.cc: g++ test-tanh.cc -o test-tanh-cc
./test-tanh-c 711
libc version 2.10.1
libc release stable
arg= 7.11e+02 + 7.11e+02 * i
sinh (arg) = inf + inf * i
cosh (arg) = inf + inf * i
s/c = nan + nan * i (should be 1 + 0 * i)
tanh (arg) = 0.00e+00 + 0.00e+00 * i (should be 1 + 0 * i)
./test-tanh-cc 711
libc version 2.10.1
arg (711,711)
sinh(arg) (1.64037e+308,2.55507e+308)
cosh(arg) (1.64037e+308,2.55507e+308)
s/c (1,0)
tanh(arg) (1,4.93127e-618)
test-tanh.c
#include
#include
#include
#define _GNU_SOURCE
#include
int main (int argc, char **argv)
{
long double complex arg = 1 + _Complex_I;
long double complex s, c, r, t;
long mult;
if (argc == 2) {
mult = atol(argv[1]);
} else {
mult = 5L;
}
arg *= mult;
s = csinh (arg);
c = ccosh (arg);
r = s / c;
t = ctanh (arg);
printf ("libc version %s\n", gnu_get_libc_version ());
printf ("libc release %s\n", gnu_get_libc_release ());
printf ("arg= %e + %e * i\n", creal (arg), cimag
(arg));
printf ("sinh (arg) = %e + %e * i\n",
creal (s), cimag (s));
printf ("cosh (arg) = %e + %e * i\n",
creal (c), cimag (c));
printf("s/c = %e + %e * i (should be 1 + 0 * i)\n",
creal (r), cimag (r));
printf ("tanh (arg) = %e + %e * i (should be 1 + 0 * i)\n",
creal (t), cimag (t));
return 0;
}
test-tanh.c
#include
#include
#include
#include // atol
using namespace std;
int main(int argc, char **argv) {
long mult;
if (argc == 2) {
mult = atol(argv[1]);
} else {
mult = 5L;
}
complex arg(1,1);
arg *= mult;
complex s,c,r,t;
s=sinh(arg);
c=cosh(arg);
r=s/c;
t=tanh(arg);
cout << "libc version " << gnu_get_libc_version() << endl;
cout << "arg " << arg << endl;
cout << "sinh(arg) " << s << endl;
cout << "cosh(arg) " << c << endl;
cout << "s/c " << r << endl;
cout << "tanh(arg) " << t << endl;
return 0;
}
--
Summary: Tanh Returns Incorrect Value
Product: gcc
Version: 4.4.1
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: libstdc++
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: tomdean at speakeasy dot org
GCC build triplet: Ubuntu 9.10 Distribution
GCC host triplet: ubuntu 9.10
GCC target triplet: i686
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=43802
[Bug libstdc++/43802] Tanh Returns Incorrect Value
--- Comment #3 from tomdean at speakeasy dot org 2010-04-19 23:33 ---
Subject: Re: Tanh Returns Incorrect Value
On Mon, 2010-04-19 at 22:56 +, pinskia at gcc dot gnu dot org wrote:
>
> --- Comment #1 from pinskia at gcc dot gnu dot org 2010-04-19 22:56
> ---
> That is because you are using the double version of ctanh rather than the long
> double version for the C case.
>
> Changing the C testcase to use ctanhl and you get the same value for the C/C++
> cases.
>
>
The problem remains.
I used a long double to try to demonstrate overflow.
Changing all the code to use ctanhl, csinhl, ccoshl, only changed the
point at wiich overflow occurs and ctanhl() returns an incorrect value.
Note: limit(tanh(n*(1+I)),n=infinity) == 1
Therefore, if the argument to ctanh is not nan or inf, ctanh should
return 1 at and above overflow for the type.
I only calculated s/c, demonstrating the overflow to another group,
octave.
Sorry for sending the wrong code, which distracted from the problem.
Here is the code I use for the test. As mult increases, you can see the
overflow happening.
>./test-tanh-c 5678
libc version 2.10.1
libc release stable
arg= 5678.00 + 5678.00 * i
sinh (arg) = -1.745126e+2465 + -3.818036e+2465 * i
cosh (arg) = -1.745126e+2465 + -3.818036e+2465 * i
s/c = 1.00 + -0.00 * i
tanh (arg) = 1.00 + 0.00 * i
>./teat-tanh-c 5679
libc version 2.10.1
libc release stable
arg= 5679.00 + 5679.00 * i
sinh (arg) = 6.170148e+2465 + -9.599250e+2465 * i
cosh (arg) = 6.170148e+2465 + -9.599250e+2465 * i
s/c = 1.00 + -0.00 * i
tanh (arg) = nan + -0.00 * i < OVERFLOW???
#include
#include
#include
#define _GNU_SOURCE
#include
int main (int argc, char **argv)
{
long double complex arg = 1 + _Complex_I;
long double complex s, c, r, t;
long mult;
if (argc == 2) {
mult = atol(argv[1]);
} else {
mult = 5L;
}
arg *= (long double)mult;
s = csinhl (arg);
c = ccoshl (arg);
r = s / c;
t = ctanhl (arg);
printf ("libc version %s\n", gnu_get_libc_version ());
printf ("libc release %s\n", gnu_get_libc_release ());
printf ("arg= %Lf + %Lf * i\n", creall (arg), cimagl
(arg));
printf ("sinh (arg) = %Le + %Le * i\n",
creall (s), cimagl (s));
printf ("cosh (arg) = %Le + %Le * i\n",
creall (c), cimagl (c));
printf("s/c = %Lf + %Lf * i\n",
creall (r), cimagl (r));
printf ("tanh (arg) = %Lf + %Lf * i\n",
creall (t), cimagl (t));
return 0;
}
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=43802
[Bug libstdc++/43802] Tanh Returns Incorrect Value
--- Comment #5 from tomdean at speakeasy dot org 2010-04-20 01:37 ---
Subject: Re: Tanh Returns Incorrect Value
On Mon, 2010-04-19 at 23:57 +, pinskia at gcc dot gnu dot org wrote:
>
> --- Comment #4 from pinskia at gcc dot gnu dot org 2010-04-19 23:57
> ---
> And tanhl is a GCC issue how? It is part of glibc and not GCC so please
> report
> it to them.
> Plus it works for me with:
> libc version 2.7
> libc release stable
>
> On Debian 5.0.
Sorry, I crossed wires when replying to your previous email. I have two
problems, one with libc and the other with libstdc++.
and, in gcc-4.4.1/libstdc++-v3/include/std/complex
template
inline complex<_Tp>
__complex_tanh(const complex<_Tp>& __z)
{ return std::sinh(__z) / std::cosh(__z); }
...
template
inline complex<_Tp>
__complex_sinh(const complex<_Tp>& __z)
{
const _Tp __x = __z.real();
const _Tp __y = __z.imag();
return complex<_Tp>(sinh(__x) * cos(__y), cosh(__x) * sin(__y));
}
...
template
inline complex<_Tp>
__complex_cosh(const complex<_Tp>& __z)
{
const _Tp __x = __z.real();
const _Tp __y = __z.imag();
return complex<_Tp>(cosh(__x) * cos(__y), sinh(__x) * sin(__y));
}
And, these all seem to resolve to gcc builtin's, like
__builtin_coshl(__x);, etc.
Is this true?
The code below was compiled with: g++ test-tanh.cc -o test-tanh-cc
>test-tanh-cc
arg (5,5)
sinh(arg) (-inf,-inf)
cosh(arg) (-inf,-inf)
s/c (nan,nan)
tanh(arg) (nan,0)
>./test-tanh-cc 5678
arg (5678,5678)
sinh(arg) (-1.74513e+2465,-3.81804e+2465)
cosh(arg) (-1.74513e+2465,-3.81804e+2465)
s/c (1,-0)
tanh(arg) (1,2.14543e-4932)
>./text-tanh-cc 5679
arg (5679,5679)
sinh(arg) (6.17015e+2465,-9.59925e+2465)
cosh(arg) (6.17015e+2465,-9.59925e+2465)
s/c (1,-0)
tanh(arg) (nan,-0)
> cat test-tanh.cc
#include
#include
#include
#include // atol
using namespace std;
int main(int argc, char **argv) {
long mult;
if (argc == 2) {
mult = atol(argv[1]);
} else {
mult = 5L;
}
complex arg(1,1);
arg *= mult;
complex s,c,r,t;
s=sinh(arg);
c=cosh(arg);
r=s/c;
t=tanh(arg);
cout << "libc version " << gnu_get_libc_version() << endl;
cout << "arg " << arg << endl;
cout << "sinh(arg) " << s << endl;
cout << "cosh(arg) " << c << endl;
cout << "s/c " << r << endl;
cout << "tanh(arg) " << t << endl;
return 0;
}
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=43802
[Bug libstdc++/43802] Tanh Returns Incorrect Value
--- Comment #7 from tomdean at speakeasy dot org 2010-04-20 02:29 --- Subject: Re: Tanh Returns Incorrect Value On Tue, 2010-04-20 at 01:58 +, pinskia at gcc dot gnu dot org wrote: Thanks for clearing that up. Please close the bug report. I filed a bug with glibc. Thanks. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=43802
