Search the user's own runtime directory before the system library

The reference desktop's only system ONNX Runtime is Arch's
onnxruntime-opt-cuda: 1.29, built without TensorRT and against cuDNN 8
on a cuDNN 9 machine. The probe rejects both providers correctly and
the app runs on the CPU provider, which is right and not what anyone
wants. runtime/ beside the models is now searched ahead of /usr/lib,
tools/fetch-desktop-runtime.sh fills it with the four libraries from
the current onnxruntime-gpu wheel (cuDNN 9, TensorRT 10), and the
About caption lists every rung that lost and why, not only the first.
Verified: the app selects TensorRT from that directory with no
environment variable set.

core/dr-inference-engine/src/lib.rs

@@ -155,6 +155,8 @@ pub struct Status {
     /// Engines compiled and engines wanted, for a compiling rung; `(0, 0)`
     /// otherwise.
     pub engines: (usize, usize),
+    /// Every rung above the selected one that was tried, and why it lost.
+    pub failed: Vec<(Rung, String)>,
 }
 
 impl Status {
@@ -399,6 +401,7 @@ pub fn status() -> Status {
         runtime: api::runtime(),
         rung,
         reason: s.cache.reason.clone(),
+        failed: s.cache.failed.clone(),
         probing: s.probing,
         engines: if rung.compiles() {
             (s.cache.compiled.len(), s.wanted)