According to the HAL specs (http://gitweb.freedesktop.org/?p=hal.git;a=blob_plain;f=doc/spec/hal-spec.html#spec-device-info) should be able to handle indirections in copy_property in both <merge>, <append> and <prepend>-nodes, however only <merge> handles indirections right. A simple .fdi file which illustrates the problem will be posted shortly. I've looked in the lastest source from the upstream repository, but it doesn't seem like they have fixed to problem yet either.
Created attachment 8579 [details] Simple test case which illustrates the bug This simple test case illustrates the bug with indirections in <append> and <prepend>. Simply put this file in /etc/hal/fdi/policy and plug in some removable storage-device (e.g an USB-key). If indirections in <append> and <prepend> had worked, volume.policy.desired_mount_point should be "usb_disk" for a USB-key.
Created attachment 8580 [details] [review] Fix for handling indirections in <append> and <prepend> This patch fixes the bug by correctly handling indirections in copy_property in <append> and <prepend>. It also simplifies the code a bit by merging the append and prepend handling code, as they are nearly identical anyway. The patch has been made with sys-apps/hal-0.5.7-r3 but also works with sys-hal-0.5.7.1-r1 without any changes.
(In reply to comment #2) > The patch has been made with sys-apps/hal-0.5.7-r3 but also works with > sys-hal-0.5.7.1-r1 without any changes. Ehm, I accidently used Gentoo-specific package names. It simple means hal-0.5.7 and hal-0.5.7.1.
Created attachment 8581 [details] [review] Fix for handling indirections in <append> and <prepend> (hal-0.5.8.1) This is a quick port of the patch to hal-0.5.8.1. It hasn't been as extensivly tested as the 0.5.7.1 patch, but it should work AFAIK.
I submitted some patches for fdi directives which cover also this bug.
fixed in git HEAD now.
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.