Try tracing handle_page_fault

JOURNAL.md

@@ -1119,6 +1119,9 @@ In today's meeting, Jonnatan suggests test the memtool program with all-in-order
 configuration. I should also try to reproduce this hang with a "production"
 bitstream (from master).
 
+Also, Xavi reports that the memtool chain test continued to run until the memory
+was exhausted using an old bistream.
+
 Let's do the quick CSR test first, and then we go back to the ftrace testing,
 which will take more time.
 
@@ -1164,3 +1167,6 @@ more slow.
 
 Interesting [article](https://wiki.osdev.org/RISC-V_Bare_Bones) on how to write
 a simple bootrom that outputs some ASCII text into the console.
+
+Regarding the `do_page_fault` filter, it seems that is no available. I need to
+pick one from `available_filter_functions` instead. Let's try `handle_page_fault`.

lagarto-ox.nix

@@ -88,7 +88,15 @@
 
         export PATH=${config.system.build.extraUtils}/bin
 
-        csrtool all-in-order
+        mkdir -p /tracing
+        mount -t tracefs nodev /tracing
+
+        cd /tracing
+        echo nop > current_tracer
+        echo 100 > max_graph_depth
+        echo handle_page_fault > set_graph_function
+        echo function_graph > current_tracer
+        cd /
 
         memtool fill $((512 * 1024 * 1024))
 
@@ -281,7 +289,7 @@
       #   tp_printk                 write ftrace events to console
       #   trace_buf_size=1M         set ftrace buffer to 1M
       #
-      setenv bootargs "root=/dev/ram0 loglevel=7 debug rw earlycon=sbi console=hvc0 rdinit=/init2 init=${init}"
+      setenv bootargs "root=/dev/ram0 loglevel=7 debug rw earlycon=sbi console=hvc0 tp_printk trace_buf_size=1M rdinit=/init2 init=${init}"
 
       EOF