Image with null bytes created on empty floppy drive

[bitsgalore]

Tested with Windows 10, running in VM.

Expected behavior:

• Ipmlab detects floppy drive is empty

Actual behavior:

• Ipmlab goes ahead, calls Aaru. Results in image with only null bytes.

This ONLY happens while floppy drive is connected via write blocker. Interestingly, the size of the empty image file always seems to correspond to the size of the most recently processed disk. Looks like the write blocker generates an empty buffer based on most recently attached medium, and exposes that to the OS.

It is possible to identify this from the log file created by Aaru, e.g. ???cic.xml lists total number of blocks:

<?xml version="1.0"?>
<CICMMetadata xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xsd="http://www.w3.org/2001/XMLSchema">
  <BlockMedia>
    <Image format="Raw disk image (sector by sector copy)" offset="0">270afede-23c0-11ed-b708-0800272c26ff.img</Image>
    <Size>1474560</Size>
    <Manufacturer>NEC</Manufacturer>
    <Model>USB UF000x</Model>
    <Interface>USB</Interface>
    <PhysicalBlockSize>512</PhysicalBlockSize>
    <LogicalBlockSize>512</LogicalBlockSize>
    <LogicalBlocks>2880</LogicalBlocks>

Then in ???resume.xml:

<?xml version="1.0"?>
<DicResume xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xsd="http://www.w3.org/2001/XMLSchema">
  <BadBlocks>
    <Block>0</Block>
    <Block>1</Block>
    <Block>2</Block>
                ::
                ::
    <Block>2877</Block>
    <Block>2878</Block>
    <Block>2879</Block>
  </BadBlocks>

While testing without a write blocker, I noticed that Ipmlab was often unable to detect floppy after 1st attempt at reading with empty floppy drive. Could be virtualisation issue, as when this happened I was also unable to access the drive from the command line (directory listing).

[bitsgalore]

Update: this also happens on Linux. Without write blocker behavior is as expected (Ipmlab detects floppy drive is empty and prompts user).

[bitsgalore]

Most likely duplicate of:

https://github.com/KBNLresearch/ipmlab/issues/26

Closing this one.