On Mon, Jan 8, 2018 at 1:18 PM, Mason <slash.tmp@xxxxxxx> wrote:
> (Could you CC me on your reply? I'm not subscribed to the ML.)
>
I'm normally using Gmane :-) But I can do email.
OK. The proper (obvious?) fix appears to bail from read_kickstart()
> if we were not able to actually read the kickstart file.
>
> Can you try the following patch, and run the program using valgrind?
>
> diff --git a/src/memory.cpp b/src/memory.cpp
> index b91d6f224f2e..ec2990c31ba4 100644
> --- a/src/memory.cpp
> +++ b/src/memory.cpp
> @@ -578,6 +578,8 @@ static int read_kickstart (struct zfile *f, uae_u8
> *mem, int size, int dochecksu
> }
> oldpos = zfile_ftell (f);
> i = zfile_fread (buffer, 1, 11, f);
> + if (i < 11) /* Unable to read kickstart file */
> + return 0;
> if (!memcmp(buffer, "KICK", 4)) {
> zfile_fseek (f, 512, SEEK_SET);
> kickdisk = 1;
>
>
>
Seems to work just fine. No output from valgrind that concerns amiberry.
