Re: [Libguestfs] [libnbd PATCH v2 5/6] block_status: Fix assertion on bad 64-bit block status reply

2023-09-22 Thread Laszlo Ersek
On 9/21/23 22:58, Eric Blake wrote: > If a server replies to a block status command with an invalid count in > NBD_REPLY_TYPE_BLOCK_STATUS_EXT, we correctly detect the server's > error, but fail to mark that we've consumed enough data off the wire > to resync back to the server's next reply.

Re: [Libguestfs] [libnbd PATCH v2 5/6] block_status: Fix assertion on bad 64-bit block status reply

2023-09-21 Thread Richard W.M. Jones
On Thu, Sep 21, 2023 at 03:58:04PM -0500, Eric Blake wrote: > If a server replies to a block status command with an invalid count in > NBD_REPLY_TYPE_BLOCK_STATUS_EXT, we correctly detect the server's > error, but fail to mark that we've consumed enough data off the wire > to resync back to the

[Libguestfs] [libnbd PATCH v2 5/6] block_status: Fix assertion on bad 64-bit block status reply

2023-09-21 Thread Eric Blake
If a server replies to a block status command with an invalid count in NBD_REPLY_TYPE_BLOCK_STATUS_EXT, we correctly detect the server's error, but fail to mark that we've consumed enough data off the wire to resync back to the server's next reply. Rich's fuzzing run initially found this, but I