It is fact that ld.so has libc statially linked into it.
Normally we wouldn't need ld.so functionality once the program is
finalyl loaded, but with the next few patches, we will have dlopen which
will reuse the same ld.so functionality.
The problem is that it seams that huge part of the code is possible not
referntially transparent. That is, it is not impossible that some of the
functions have internals states. So when using the struct Linker that is
initialized by ld.so's copy of libc. we must access it using the same
copy even if both copies are identical.
For example in dlopen if you do linker.load_library(..). That would
segfault because it is using the function from libc not ld.so
So I don't truly undestand why should this be needed, but after long
hours of being stuck I thought maybe.. maybe that is the issue and
indeed it turned out to be so.
I attempted fixing this issue before at 43fbaf99. Although it did work,
it worked wrong, and it was just consistently working (but in wrong way)
until it didn't.
Since this is (hopefully) the real fix, I will try to explain exactly
what is going on.
This is explaination by example:
our TLS is memory of size 0x1000 starting at 0x7ffff6c50000,
but the real size is 0x000068 so we have padding stored at master.offset
= 0xf98
Now our symbol looks as follows
Offset Type Sym. Value Name
000000432b20 R_X86_64_DTPOFF64 0000000000000058 errno
The old code did 0x7ffff6c50000 + 0xf98 + 000000432b20 which is
obviosly overflowing the memory and wrong.
The right way 0x7ffff6c50000 + 0xf98 + 0000000000000058.
THe Tls base part and offset are added at __tls_get_addr function.
What is left is storing the 0x58 at the relocation address. The problem
is that we don't have 0x58, but we have (binary base + 0x58) in global
symbol table and binary base so what we store is the (binarybase + 0x58
- binary base).
I hope this does turn out to be wrong.
When a byte-oriented stream function touches a stream, that stream
should be set to byte-oriented mode if it hasn't been set yet. If
it has been set, the opertion should only succeed if the stream is
already in byte-oriented mode.
Signed-off-by: Wren Turkal <wt@penguintechs.org>
This function is used to set the orientation of a stream to either
byte-oriented or wchar-oriented.
More info on this function is here:
https://man7.org/linux/man-pages/man3/fwide.3p.html
This implementation only impmlemnts the manual switching and does
not yet guard against using a byte-oriented stream with wchar
functions and vice versa. Those step will come in additional
commits.
Signed-off-by: Wren Turkal <wt@penguintechs.org>
The problem here was that we alway added the base address, and we
assumed that all addresses we access are relative but this is not the
case in case of non pie binaries. The issue is that all addresses were
base+offset. so if we added the base again it will ofcourse generate
wrong address.
This was triggered by gcc for some reason It included sys/types.h and
assumed sys/select.h to be there. And that seams to be the case in musl.
The problem with relibc here is that sys/types.h is are part of relibc
"include/*.h" files, while sys/select.h is generated by cbindgen. That
makes it impossible to #include select.h in types.h epsecially that
there are files like fcntl.c that uses types.h. They would complain
about missing headers. I fixed this by renaming sys/types.h to
sys/types_internal.h and then generating types.h using cbindgen as well
except for that. however fcntl and dlmalloc can include types_internal
instead of types.h
The problem here is that _Bool type is not defined in C++ yet this file
is using it. That leads to issues when compiling gcc. I borrowed the
same techniques used in other stdbool.h
oddcoder