diff --git a/chapter08/coreutils.xml b/chapter08/coreutils.xml
index 94c276360..68ff3b356 100644
--- a/chapter08/coreutils.xml
+++ b/chapter08/coreutils.xml
@@ -123,20 +123,18 @@ FORCE_UNSAFE_CONFIGURE=1 ./configure \
 
 <screen><userinput remap="test">chown -R tester . </userinput></screen>
 
-    <para>Now run the tests:</para>
+    <!-- To editors: in jhalfs the standard input is always piped to "make"
+         so the test failures are not reproducible in jhalfs even if
+         running the build in a PTY.  Don't use jhalfs to test if we can
+         remove < /dev/null.  -->
+    <para>Now run the tests (using <filename>/dev/null</filename> for the
+    standard input, or two tests may be broken if building LFS in a
+    graphical terminal or a session in SSH or GNU Screen etc. because the
+    standard input is connected to a PTY from host distro, and the device
+    node for such a PTY cannot be accessed from the LFS chroot
+    environment):</para>
 
-<screen><userinput remap="test">su tester -c "PATH=$PATH make RUN_EXPENSIVE_TESTS=yes check"</userinput></screen>
-
-<!-- Sometimes skipped, but if not it may fail.
-     Please DO NOT remove or comment out this unless we can fully
-     understand it!
-    <para>The test-getlogin test may fail in the LFS chroot environment.</para>
-    20240226: Commented out as it seems the upstream has root-caused and
-    fixed the issue.  I've checked the fix is in Coreutils 9.4 release
-    tarball:
-
-    https://lists.gnu.org/archive/html/bug-gnulib/2022-06/msg00079.html
--->
+<screen><userinput remap="test">su tester -c "PATH=$PATH make RUN_EXPENSIVE_TESTS=yes check" &lt; /dev/null</userinput></screen>
 
     <para>Remove the temporary group:</para>
 
diff --git a/chapter08/glibc.xml b/chapter08/glibc.xml
index adbbcc480..1d217bfa9 100644
--- a/chapter08/glibc.xml
+++ b/chapter08/glibc.xml
@@ -178,9 +178,7 @@ esac</userinput></screen>
         relatively slow and/or running the test suite with multiple
         parallel make jobs).  These tests can be identified with:</para>
 
-        <!-- TODO: Using nodump for freeze.  Change it to role="test" after
-             12.1 release so jhalfs can list these in the log.  -->
-        <screen role="nodump"><userinput>grep "Timed out" -l $(find -name \*.out)</userinput></screen>
+        <screen><userinput remap="test">grep "Timed out" $(find -name \*.out)</userinput></screen>
 
         <para>It's possible to re-run a single test with enlarged timeout
         with
diff --git a/chapter08/procps.xml b/chapter08/procps.xml
index 3f229c713..b592f75c2 100644
--- a/chapter08/procps.xml
+++ b/chapter08/procps.xml
@@ -70,22 +70,17 @@
 <screen revision='sysv'><userinput remap="make">make</userinput></screen>
 <screen revision='systemd'><userinput remap="make">make src_w_LDADD='$(LDADD) -lsystemd'</userinput></screen>
 
-    <para> To run the test suite, run:</para>
+    <!-- To editors: "pmap {X,XX} with unreachable process" tests don't work
+         as root.  They may seem to pass due to some Tcl/Expect/DejaGNU
+         internal hazards with some "luck" though.  -->
+    <para>To run the test suite, run:</para>
 
-<screen><userinput remap="test">make -k check</userinput></screen>
+<screen><userinput remap="test">su tester -c "PATH=$PATH make check"</userinput></screen>
 
     <para>One test named
     <literal>ps with output flag bsdtime,cputime,etime,etimes</literal>
     is known to fail if the host kernel is not built with
-    <literal>CONFIG_BSD_PROCESS_ACCT</literal> enabled.
-    <!-- These two tests just do not work as root but they seem to pass
-         in most time due to some internal hazards of Tcl/Expect/DejaGNU
-         framework.  However the failures have been really observed.
-         TODO: run the test suite as tester after the freeze ends.  -->
-    Two tests named
-    <literal>pmap X with unreachable process</literal> and
-    <literal>pmap XX with unreachable process</literal> are known to
-    fail occasionally.</para>
+    <literal>CONFIG_BSD_PROCESS_ACCT</literal> enabled.</para>
 
     <para>Install the package:</para>