[Bug ld/30254] iostream crashes with ld linker for windows targets
https://sourceware.org/bugzilla/show_bug.cgi?id=30254 cqwrteur changed: What|Removed |Added CC||jakub at redhat dot com -- You are receiving this mail because: You are on the CC list for the bug.
[Bug ld/30375] C++ std::cout's rdbuf gives ridiculous address on windows with dynamic linking with ld
https://sourceware.org/bugzilla/show_bug.cgi?id=30375 cqwrteur changed: What|Removed |Added Resolution|--- |DUPLICATE Status|UNCONFIRMED |RESOLVED --- Comment #2 from cqwrteur --- The bug fixed *** This bug has been marked as a duplicate of bug 29973 *** -- You are receiving this mail because: You are on the CC list for the bug.
[Bug ld/30254] iostream crashes with ld linker for windows targets
https://sourceware.org/bugzilla/show_bug.cgi?id=30254 cqwrteur changed: What|Removed |Added Status|UNCONFIRMED |RESOLVED Resolution|--- |FIXED --- Comment #18 from cqwrteur --- (In reply to cqwrteur from comment #17) > Created attachment 14898 [details] > English translation https://sourceware.org/bugzilla/attachment.cgi?id=14656 Looks like I applied the wrong patch the dev proposed and that is how it causes the problem. Closed Thank you the old issue fixed now. -- You are receiving this mail because: You are on the CC list for the bug.
[Bug ld/29973] x86_64-w64-mingw32-g++ ld: helloworld.exe:.rdata_r: section below image base for windows
https://sourceware.org/bugzilla/show_bug.cgi?id=29973 --- Comment #42 from cqwrteur --- *** Bug 30375 has been marked as a duplicate of this bug. *** -- You are receiving this mail because: You are on the CC list for the bug.
[Bug ld/30254] iostream crashes with ld linker for windows targets
https://sourceware.org/bugzilla/show_bug.cgi?id=30254 cqwrteur changed: What|Removed |Added CC||amodra at gmail dot com -- You are receiving this mail because: You are on the CC list for the bug.
[Bug ld/30254] iostream crashes with ld linker for windows targets
https://sourceware.org/bugzilla/show_bug.cgi?id=30254 --- Comment #10 from cqwrteur --- Created attachment 14895 --> https://sourceware.org/bugzilla/attachment.cgi?id=14895=edit ld no link -- You are receiving this mail because: You are on the CC list for the bug.
[Bug ld/30254] iostream crashes with ld linker for windows targets
https://sourceware.org/bugzilla/show_bug.cgi?id=30254 --- Comment #12 from cqwrteur --- Ida pro result for GNU ld. I do not see jmp instruction here. .text:000140001540 ; === S U B R O U T I N E === .text:000140001540 .text:000140001540 ; Attributes: bp-based frame .text:000140001540 .text:000140001540 ; int __cdecl main() .text:000140001540 public main .text:000140001540 main proc near ; CODE XREF: __tmainCRTStartup+169↑p .text:000140001540 ; DATA XREF: .pdata:000140006084↓o .text:000140001540 55pushrbp .text:000140001541 48 89 E5 mov rbp, rsp .text:000140001544 48 83 EC 20 sub rsp, 20h .text:000140001548 ; #line 4 .text:000140001548 E8 F3 00 00 00call__main .text:000140001548 .text:00014000154D ; #line 5 .text:00014000154D 48 8D 05 FC 2A 00 00 lea rax, aHelloWorld ; "Hello World\n" .text:000140001554 48 89 C2 mov rdx, rax .text:000140001557 48 8B 05 72 2E 00 00 mov rax, cs:_refptr__ZSt4cout .text:00014000155E 48 89 C1 mov rcx, rax .text:000140001561 E8 1A 00 00 00call _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc ; std::operator<<>(std::ostream &,char const*) .text:000140001561 .text:000140001566 ; #line 6 .text:000140001566 B8 00 00 00 00mov eax, 0 .text:00014000156B 48 83 C4 20 add rsp, 20h .text:00014000156F 5Dpop rbp .text:000140001570 C3retn .text:000140001570 .text:000140001570 main endp .text:000140001570 .text:000140001570 ; --- .text:000140001571 ; #line 6 .text:000140001571 algn_140001571: ; DATA XREF: .pdata:000140006084↓o .text:000140001571 90 90 90 90 90 90 90 90 90 90+align 20h .text:000140001580 ; [0006 BYTES: COLLAPSED FUNCTION std::operator<<>(std::ostream &,char const*). PRESS CTRL-NUMPAD+ TO EXPAND] .text:000140001586 90 90 align 8 .text:000140001588 ; [0006 BYTES: COLLAPSED FUNCTION std::ios_base_library_init(void). PRESS CTRL-NUMPAD+ TO EXPAND] .text:00014000158E 90 90 align 10h .text:000140001590 ; #line 21 .text:000140001590 .text:000140001590 ; === S U B R O U T I N E === .text:000140001590 -- You are receiving this mail because: You are on the CC list for the bug.
[Bug ld/30254] iostream crashes with ld linker for windows targets
https://sourceware.org/bugzilla/show_bug.cgi?id=30254 --- Comment #13 from cqwrteur --- .text:000140001588 ; === S U B R O U T I N E === .text:000140001588 .text:000140001588 ; Attributes: thunk .text:000140001588 .text:000140001588 ; _QWORD std::ios_base_library_init(void) .text:000140001588 public _ZSt21ios_base_library_initv .text:000140001588 _ZSt21ios_base_library_initv proc near .text:000140001588 FF 25 62 7D 00 00 jmp cs:__imp__ZSt21ios_base_library_initv ; std::ios_base_library_init(void) .text:000140001588 .text:000140001588 _ZSt21ios_base_library_initv endp Should this function be in pdata section?? -- You are receiving this mail because: You are on the CC list for the bug.
[Bug ld/30254] iostream crashes with ld linker for windows targets
https://sourceware.org/bugzilla/show_bug.cgi?id=30254 --- Comment #8 from cqwrteur --- Created attachment 14893 --> https://sourceware.org/bugzilla/attachment.cgi?id=14893=edit lld has link from text section that jmps here while ld does not lld has link from text section that jmps here while ld does not -- You are receiving this mail because: You are on the CC list for the bug.
[Bug ld/30254] iostream crashes with ld linker for windows targets
https://sourceware.org/bugzilla/show_bug.cgi?id=30254 --- Comment #9 from cqwrteur --- Created attachment 14894 --> https://sourceware.org/bugzilla/attachment.cgi?id=14894=edit ld no link ld no link 1 -- You are receiving this mail because: You are on the CC list for the bug.
[Bug ld/30254] iostream crashes with ld linker for windows targets
https://sourceware.org/bugzilla/show_bug.cgi?id=30254 --- Comment #11 from cqwrteur --- Here is the disassembly result from IDA pro. .text:000140001540 ; === S U B R O U T I N E === .text:000140001540 .text:000140001540 ; Attributes: bp-based frame .text:000140001540 .text:000140001540 ; int __cdecl main() .text:000140001540 main proc near ; CODE XREF: __tmainCRTStartup+169↑p .text:000140001540 ; DATA XREF: .pdata:000140007084↓o .text:000140001540 55pushrbp .text:000140001541 48 89 E5 mov rbp, rsp .text:000140001544 48 83 EC 20 sub rsp, 20h .text:000140001548 ; #line 4 .text:000140001548 E8 F3 00 00 00call__main .text:000140001548 .text:00014000154D ; #line 5 .text:00014000154D 48 8D 05 4C 1C 00 00 lea rax, aHelloWorld ; "Hello World\n" .text:000140001554 48 89 C2 mov rdx, rax .text:000140001557 48 8B 05 52 1C 00 00 mov rax, cs:off_1400031B0 .text:00014000155E 48 89 C1 mov rcx, rax .text:000140001561 E8 1A 00 00 00call _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc ; std::operator<<>(std::ostream &,char const*) .text:000140001561 .text:000140001566 ; #line 6 .text:000140001566 B8 00 00 00 00mov eax, 0 .text:00014000156B 48 83 C4 20 add rsp, 20h .text:00014000156F 5Dpop rbp .text:000140001570 C3retn .text:000140001570 .text:000140001570 main endp .text:000140001570 .text:000140001570 ; --- .text:000140001571 ; #line 6 .text:000140001571 algn_140001571: ; DATA XREF: .pdata:000140007084↓o .text:000140001571 90 90 90 90 90 90 90 90 90 90+align 20h .text:000140001580 ; [0006 BYTES: COLLAPSED FUNCTION std::operator<<>(std::ostream &,char const*). PRESS CTRL-NUMPAD+ TO EXPAND] .text:000140001586 90 90 align 8 .text:000140001588 FF 25 D2 28 00 00 jmp cs:_ZSt21ios_base_library_initv ; std::ios_base_library_init(void) .text:000140001588 .text:000140001588 ; --- .text:00014000158E 90 90 align 10h .text:000140001590 ; #line 21 We can see .pdata:000140007084 has a link to std::ios_base_library_init in libstdc++-6.dll for LLVM lld. However, GNU ld does not do the right thing at here. It looks like a serious bug in the linker. -- You are receiving this mail because: You are on the CC list for the bug.
[Bug ld/30254] iostream crashes with ld linker for windows targets
https://sourceware.org/bugzilla/show_bug.cgi?id=30254
--- Comment #14 from cqwrteur ---
Some one tries this program:
#include
#include
#include
#include
typedef void* (__cdecl* fn_type_t)(void* self, char const* string);
int main()
{
MessageBoxA(nullptr,"pause",nullptr,0);
auto hModule = GetModuleHandleA("libstdc++-6.dll");
auto cout = reinterpret_cast(GetProcAddress(hModule,
"_ZSt4cout"));
auto fn = reinterpret_cast(GetProcAddress(hModule,
"_ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc"));
fn(cout, "Hello World!\n");
std::cout << "Hello World!\n";
return 0;
}
He found std::cout binds to the wrong address for ld.
Here is a variable inside the PE file.
void * address = ::cout;
Now, we want to make this = address, but GNU LD gives you this =
This pointer is incorrect,
so it will definitely result in an error.
--
You are receiving this mail because:
You are on the CC list for the bug.
[Bug ld/30254] iostream crashes with ld linker for windows targets
https://sourceware.org/bugzilla/show_bug.cgi?id=30254 --- Comment #15 from cqwrteur --- Created attachment 14896 --> https://sourceware.org/bugzilla/attachment.cgi?id=14896=edit He found that ld binds std::cout to the address of std::cout in the dll Why is this .refptr?? It should be just bound to _ZSt4cout -- You are receiving this mail because: You are on the CC list for the bug.
[Bug ld/30254] iostream crashes with ld linker for windows targets
https://sourceware.org/bugzilla/show_bug.cgi?id=30254 --- Comment #7 from cqwrteur --- Created attachment 14892 --> https://sourceware.org/bugzilla/attachment.cgi?id=14892=edit executable -- You are receiving this mail because: You are on the CC list for the bug.
[Bug ld/30254] iostream crashes with ld linker for windows targets
https://sourceware.org/bugzilla/show_bug.cgi?id=30254
--- Comment #6 from cqwrteur ---
Created attachment 14891
--> https://sourceware.org/bugzilla/attachment.cgi?id=14891=edit
ld missing import of ios_base_library_init
I use lld and it imports one more symbol than ld.
I guess it is the reason since libstdc++ does this:
#ifdef _GLIBCXX_SYMVER_GNU
#pragma GCC diagnostic ignored "-Wattribute-alias"
void ios_base_library_init (void)
__attribute__((alias ("_ZNSt8ios_base4InitC1Ev")));
#endif
https://github.com/gcc-mirror/gcc/blob/master/libstdc%2B%2B-v3/src/c%2B%2B98/ios_init.cc
--
You are receiving this mail because:
You are on the CC list for the bug.
[Bug ld/30254] iostream crashes with ld linker for windows targets
https://sourceware.org/bugzilla/show_bug.cgi?id=30254 --- Comment #16 from cqwrteur --- Created attachment 14897 --> https://sourceware.org/bugzilla/attachment.cgi?id=14897=edit Someone has found the problem here Someone has found the problem here. std::cout does not bind correctly for ld -- You are receiving this mail because: You are on the CC list for the bug.
[Bug ld/30254] iostream crashes with ld linker for windows targets
https://sourceware.org/bugzilla/show_bug.cgi?id=30254 --- Comment #17 from cqwrteur --- Created attachment 14898 --> https://sourceware.org/bugzilla/attachment.cgi?id=14898=edit English translation -- You are receiving this mail because: You are on the CC list for the bug.
Issue 59148 in oss-fuzz: binutils:fuzz_addr2line: Heap-buffer-overflow in bfd_getl32
Updates: Labels: -restrict-view-commit Comment #3 on issue 59148 by sheriffbot: binutils:fuzz_addr2line: Heap-buffer-overflow in bfd_getl32 https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=59148#c3 This bug has been fixed. It has been opened to the public. - Your friendly Sheriffbot -- You received this message because: 1. You were specifically CC'd on the issue You may adjust your notification preferences at: https://bugs.chromium.org/hosting/settings Reply to this email to add a comment.
[Bug binutils/30479] New: bfd.info build failure on centos 8
https://sourceware.org/bugzilla/show_bug.cgi?id=30479 Bug ID: 30479 Summary: bfd.info build failure on centos 8 Product: binutils Version: unspecified Status: UNCONFIRMED Severity: normal Priority: P2 Component: binutils Assignee: unassigned at sourceware dot org Reporter: nightstrike at gmail dot com Target Milestone: --- See https://sourceware.org/bugzilla/show_bug.cgi?id=29476 for related problems. The 2.40 tar release requires makeinfo as well. So while gprof was fixed, bfd is still a problem. I am assuming that building the documentation for a release tar is not supposed to be required. In this case, I am trying it on a system that doesn't have makeinfo installed at all. -- You are receiving this mail because: You are on the CC list for the bug.
[Bug binutils/30479] bfd.info build failure on centos 8
https://sourceware.org/bugzilla/show_bug.cgi?id=30479 --- Comment #2 from nightstrike --- Renaming the included texinfo directory to test configure's ability to build texinfo shows that that doesn't even work anymore: /bin/sh: line 7: /tmp/src/binutils-2.40/_/build-x86_64-pc-linux-gnu/texinfo/makeinfo/makeinfo: No such file or directory /bin/sh: line 7: /tmp/src/binutils-2.40/_/build-x86_64-pc-linux-gnu/texinfo/makeinfo/makeinfo: No such file or directory make[5]: *** [Makefile:1564: texinfo.info] Error 127 The build-x86_64-pc-linux-gnu dir is never created, let alone anything else. -- You are receiving this mail because: You are on the CC list for the bug.
[Bug binutils/30479] bfd.info build failure on centos 8
https://sourceware.org/bugzilla/show_bug.cgi?id=30479 nightstrike changed: What|Removed |Added Version|unspecified |2.40 -- You are receiving this mail because: You are on the CC list for the bug.
[Bug binutils/30481] New: bfd version inconsistencies
https://sourceware.org/bugzilla/show_bug.cgi?id=30481 Bug ID: 30481 Summary: bfd version inconsistencies Product: binutils Version: 2.40 Status: UNCONFIRMED Severity: normal Priority: P2 Component: binutils Assignee: unassigned at sourceware dot org Reporter: nightstrike at gmail dot com Target Milestone: --- The following warnings are emitted during the build, many times each: In file included from ../../gprof/../bfd/sysdep.h:29, from ../../gprof/gprof.h:33: ../bfd/config.h:327: note: this is the location of the previous definition 327 | #define VERSION "2.40" | In file included from ../../gprof/gprof.h:41, from ../../gprof/corefile.c:22: ./gconfig.h:104: warning: "VERSION" redefined 104 | #define VERSION "2.40.00" | There is an inconsistency in naming convention regarding 2.40 and 2.40.00. -- You are receiving this mail because: You are on the CC list for the bug.
[Bug binutils/30479] bfd.info build failure on centos 8
https://sourceware.org/bugzilla/show_bug.cgi?id=30479 --- Comment #1 from nightstrike --- It looks like adding texinfo to top level to be built automatically (like isl, gmp, mpfr, mpc, dejagnu, and a bunch of other stuff) doesn't work OOTB, because the binutils tarball include a texinfo directory containing a texinfo.tex file. That invalidates all of the top level configure rules, such as "maybe-configure-texinfo", etc. -- You are receiving this mail because: You are on the CC list for the bug.
[Bug binutils/30481] bfd version inconsistencies
https://sourceware.org/bugzilla/show_bug.cgi?id=30481 Alan Modra changed: What|Removed |Added Resolution|--- |FIXED Status|UNCONFIRMED |RESOLVED --- Comment #1 from Alan Modra --- This has already been fixed on the 2.40 branch. -- You are receiving this mail because: You are on the CC list for the bug.
[Bug binutils/30479] bfd.info build failure on centos 8
https://sourceware.org/bugzilla/show_bug.cgi?id=30479 --- Comment #3 from nightstrike --- Ok, I worked around it for now by touching bfd/doc/bfd.info and gas/doc/as.info (which also exhibited the problem) in the source dir. I had to do this individually, and before each rebuild. I think the build system can be improved here, though. And certainly, texinfo should be buildable. -- You are receiving this mail because: You are on the CC list for the bug.
