> i've been reviewing open-vm-tools[1] and Russ' vmwarefs. from what i > can tell the old backdoor (low bandwidth) mechanism is still supported > and the protocol hasn't changed. there's an alternative high > bandwidth mechanism but i'm not sure if the advent of the high > bandwidth rpc was later than plan9's vmware tools (Russ?).
it happened about the same time. there is support for the high bandwidth channel -- read and write /mnt/vmware/msg. > i setup a new vm (workstation v. 6.5.1) and installed from the latest > iso image and compared it to my existing vm (4.5.3) running an older > kernel. the results from directly io to /mnt/vmware/snarf are the > same as far as i can tell. this pattern is consistent: the snarf > buffer of the Plan 9 guest can be made visible in Windows host > consistently ONLY if in Plan 9 guest, the snarf file is read before > writing to it. that is: > > cat /mnt/vmware/snarf > echo 1 > /mnt/vmware/snarf interesting. maybe that's worth making a new binary. russ
