A Bug in Disk Tainting?

[M Chen]

Hello, 

Thanks for your time, I have a question about Decaf disk tainting. 

At code: https://github.com/sycurelab/DECAF/blob/master/decaf/hw/ide/pci.c line 138: I found that Decaf calls taintcheck_chk_hdread() after pci_dma_write(); and calls taintcheck_chk_hdwrite() after pci_dma_read(). 

If I understand it correctly, the codes act as: 

    when Qemu does dma write, Decaf will check the disk taint (if any); and when Qemu does dma read, Decaf taints the disk file via taintcheck_chk_hdwrite().

However, shouldn't it be the reversed logic? That is, when Qemu does a write, Decaf will taint the target disk file during which; when Qemu does a read, Decaf will check if the file being read is tainted or not?

Or I misunderstand the behavior of Decaf disk tainting?

Thanks,

Michael Chen

[Heng Yin]

pci_dma_read means a PCI device reads from memory. In other words, it is writing into the device. 

Heng

--
You received this message because you are subscribed to the Google Groups "decaf-platform-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to decaf-platform-di...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.