Hi Brian, Andrew,
Thanks so much for the help. The issue is now resolved after
implementing your suggested changes. Apologies for asking for
assistance on such a basic question ... I was obviously able to get
away with this before (on 3.4.3), but I'm glad to see the newer
compilers are rejecting this now.
John Doe
On 1/2/07, Andrew Haley <aph@xxxxxxxxxxx> wrote:
John Doe writes:
>
> > without seeing the input that causes the error nobody will really
> > be able to tell you much.
>
> Many thanks for the response. I've pasted the relevant code below:
>
> typedef void* HANDLETYPE;
>
> ERRORTYPE VIDDEC_HandleCommand (HANDLETYPE phandle, ...)
> {
>
> VIDDEC_COMPONENT_PRIVATE* pComponentPrivate =
> (VIDDEC_COMPONENT_PRIVATE*)phandle;
> HANDLETYPE hIPC = NULL;
> ....
> ....
> (IPC_DSP_INTERFACE*)pComponentPrivate->pIPC = (IPC_DSP_INTERFACE*)hIPC;
>
> The line immediately above this text is causing the aforementioned
> "invalid lvalue in assignment" error. Would this qualify as one of
> the deprecated conditions you mentioned below?
This construct is known as an lvalue cast. It used to be supported in
gcc, but isn't any longer.
It's easy enough to fix:
pComponentPrivate->pIPC = (pIPCType*)hIPC;
Andrew.
