Segfaulting in FFI #91
Labels
Comments
|
Can confirm, caused by #90. Keeping the pub struct Library {
raw: ffi::FT_Library,
// Hold onto this because FT_Library references it internally.
_memory: Box<ffi::FT_MemoryRec>,
}
impl Library {
pub fn init() -> FtResult<Library> {
let memory = box ffi::FT_MemoryRec {
user: 0 as *mut c_void,
alloc: alloc_library,
free: free_library,
realloc: realloc_library,
};
unsafe {
let mut raw = std::ptr::null_mut();
let err = ffi::FT_New_Library(&*memory, &mut raw);
if err == ffi::FT_Err_Ok {
ffi::FT_Add_Default_Modules(raw);
Ok(Library {
raw: raw,
_memory: memory,
})
} else {
Err(FromPrimitive::from_i32(err).unwrap())
}
}
}
// ...
}It's segfaulting in other places still so I'm calling this solution a work in progress. |
|
Another backtrace (with Relevant function: FT_EXPORT_DEF( FT_Error )
FT_Done_Face( FT_Face face )
{
FT_Error error;
FT_Driver driver;
FT_Memory memory;
FT_ListNode node;
error = FT_ERR( Invalid_Face_Handle );
if ( face && face->driver )
{
face->internal->refcount--;
if ( face->internal->refcount > 0 )
error = FT_Err_Ok;
else
{
driver = face->driver;
memory = driver->root.memory;
/* find face in driver's list */
node = FT_List_Find( &driver->faces_list, face );
if ( node )
{
/* remove face object from the driver's list */
FT_List_Remove( &driver->faces_list, node );
FT_FREE( node );
/* now destroy the object proper */
destroy_face( memory, face, driver );
error = FT_Err_Ok;
}
}
}
return error;
}Not sure what might be segfaulting here. I wish I had debug symbols. Maybe I'll work on a debug build of Freetype tonight. |
|
Got Freetype with debug symbols! Things are getting interesting. Backtrace of previous comment, plus some debugging: I think |
|
Nice work @cybergeek94! 👍 |
|
I tried the Box solution and continued see crashing as you mentioned. Rc seems to work for me though: use std::rc::Rc;
//...
pub struct Library {
raw: ffi::FT_Library,
_memory: Rc<ffi::FT_MemoryRec>,
}
impl Library {
pub fn init() -> FtResult<Library> {
let memory = Rc::new(ffi::FT_MemoryRec {
user: 0 as *mut c_void,
alloc: alloc_library,
free: free_library,
realloc: realloc_library,
});
unsafe {
let mut raw = std::ptr::null_mut();
let err = ffi::FT_New_Library(memory.deref(), &mut raw);
if err == ffi::FT_Err_Ok {
ffi::FT_Add_Default_Modules(raw);
Ok(Library {
raw: raw,
_memory: memory,
})
} else {
Err(FromPrimitive::from_i32(err).unwrap())
}
}
}
// ... |
|
@jakerr Looks like you've got it! Still having trouble with You want to issue a PR or should I? |
|
Still having trouble in a multithreaded context, even though I created I have no idea what to make of this. The function pointers appear to be initialized properly, but when I try to call them, they segfault: |
|
I fixed it. It was easy. In struct FT_MemoryRec {
//...
}
unsafe impl Sync for FT_MemoryRec {}Then revert #90. No more segfaults, even in threaded contexts. Alternately, we could avoid dealing with |
|
@cybergeek94 Thanks for working on this! Reverted with #92 |
This was referenced Jan 1, 2015
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
I downloaded Freetype's source to get the body of the relevant function (since there doesn't seem to be a debug symbols package available in the Ubuntu/Mint repos ¯_(ツ)_/¯ ):
The only place a segfault looks possible is the virtual call
memory->alloc()which I think would segfault ifmemorywas null, correct?I can't figure out how to configure a debug build for Freetype or else I'd have tried to debug this further. I mean, I can configure it, but the
./configurescript is giving me weird errors.Is this possibly related to #90?
The text was updated successfully, but these errors were encountered: