On Thursday 17 February 2011 4:19:17 pm Stefan Hajnoczi wrote: > On Wed, Feb 16, 2011 at 12:23 PM, M. Mohan Kumar <mo...@in.ibm.com> wrote: > > + switch (type) { > > + case T_MKNOD: > > + request.data.type = T_MKNOD; > > + break; > > + case T_MKDIR: > > + request.data.type = T_MKDIR; > > + break; > > + case T_SYMLINK: > > + request.data.type = T_SYMLINK; > > + break; > > + case T_LINK: > > + request.data.type = T_LINK; > > + break; > > + } > > Or just request.data.type = type? > Oops, I will modify. > > + if (oldpath) { > > + request.data.oldpath_len = strlen(oldpath); > > + strcpy(request.path.old_path, oldpath); > > It's not obvious that this strcpy() is safe. >
I will change the code. > > + } > > + retval = v9fs_create_special(fs_ctx, &request, &error); > > + if (retval < 0) { > > + errno = error; > > + return 0; > > This looks suspicious. Should the return value be negative on error? Yes, it should return negative on error, I will update. > > > + char *tmp = qemu_strdup(rpath(fs_ctx, oldpath)); > > + if (tmp == NULL) { > > QEMU strdup never returns NULL. Ok. > > Stefan ---- M. Mohan Kumar