Nikolay Sivov nsivov@codeweavers.com writes:
SERVER_START_REQ( get_token_groups ) {
need_more_memory = FALSE;
TOKEN_GROUPS *groups = tokeninfo;
void *buffer;
SERVER_START_REQ( get_token_groups )
{
TOKEN_GROUPS *groups = tokeninfo;
/* reply buffer is always shorter than output one */
buffer = tokeninfolength ? RtlAllocateHeap(GetProcessHeap(), 0, tokeninfolength) : NULL;
Please avoid complex operation like heap allocations inside a server request block.