I use the C API.
I think it is not possible to get the hole call stack if continue using the
released SQLite pointer.
???:Simon Slavinslavins at bigfraud.org
???:SQLite mailing listsqlite-users at mailinglists.sqlite.org
:2016?3?4?(??)?17:50
??:Re: [sqlite] SQLite Corruption By Writing
I am debugging db corruption. After I get some corrupted db, I found that they
all corrupted by writing null data.
So, I decide to add some check and dump call stackin the source code in order
to find out who corrupts the db.
Here is the code I added in the source code.
int sqlite3CheckNullDa
sanhua.zh wrote:
> I am debugging db corruption. After I get some corrupted db, I found that
> they all corrupted by writing null data.
>
> 0x1000f8000 + 2778664 sqlite3KnownError,main.c,line 3192
> 0x1000f8000 + 2554560 unixWrite,os_unix.c,line 3335
> 0x1000f8000 + 2821984 sqlite3WalCheckpo
On 4 Mar 2016, at 8:22am, sanhua.zh wrote:
> 3. I guess it could be a problem of operating system. I work on iOS, but I
> have no any further idea.
Almost all of these problems are caused by your program doing one of these
A) Writing its own data into a pointer made by SQLite
B) Releasing a S
4 matches
Mail list logo