01.11.2014 04:50, zhanghailiang wrote:
> For qemu_chr_open_spice_vmc and qemu_chr_open_spice_port, the in-parameter
> never to be NULL, because the checks in qemu_chr_parse_spice_vmc
> and qemu_chr_parse_spice_port have ensured this.
>
> So we should check the length of the in-parameter.
The same applies here as to qemu_chr_open in patch 1/4.
But here we've one more thing:
> Signed-off-by: zhanghailiang <zhang.zhanghaili...@huawei.com>
> ---
> spice-qemu-char.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/spice-qemu-char.c b/spice-qemu-char.c
> index 8106e06..45e7d69 100644
> --- a/spice-qemu-char.c
> +++ b/spice-qemu-char.c
> @@ -290,7 +290,7 @@ CharDriverState *qemu_chr_open_spice_vmc(const char *type)
> {
> const char **psubtype = spice_server_char_device_recognized_subtypes();
>
> - if (type == NULL) {
> + if (type == NULL || strlen(type) == 0) {
> fprintf(stderr, "spice-qemu-char: missing name parameter\n");
This is 'missing TYPE parameter' not name. If we merge the check with
qemu_chr_parse_* it will go away.
Thanks,
/mjt
> print_allowed_subtypes();
> return NULL;
> @@ -315,7 +315,7 @@ CharDriverState *qemu_chr_open_spice_port(const char
> *name)
> CharDriverState *chr;
> SpiceCharDriver *s;
>
> - if (name == NULL) {
> + if (name == NULL || strlen(name) == 0) {
> fprintf(stderr, "spice-qemu-char: missing name parameter\n");
> return NULL;
> }
>
