Use (patched) staticx for Aarch64 exe builds (#355)

Author: Jongy

exe-requirements.txt

@@ -3,4 +3,11 @@ pyinstaller==4.0; platform.machine == "x86_64"
 # aarch64 requires a later version due to the use of a newer centos version.
 # see https://github.com/pyinstaller/pyinstaller/issues/5540
 pyinstaller==4.10; platform.machine == "aarch64"
-staticx==0.12.1
+# for aarch64 we build a slightly patched version
+# I tried upgrading to 0.13.6 but it crashes when the botoloader is run as non-root :/
+# I got this error in gdb:
+# (gdb) run
+# Starting program: /path/to/build/x86_64/gprofiler
+# During startup program terminated with signal SIGSEGV, Segmentation fault.
+# staying with 0.12.1 for the mean time...
+staticx==0.12.1; platform.machine == "x86_64"

pyi.Dockerfile

@@ -204,14 +204,24 @@ RUN pyinstaller pyinstaller.spec \
     && test -f build/pyinstaller/warn-pyinstaller.txt \
     && if grep 'gprofiler\.' build/pyinstaller/warn-pyinstaller.txt ; then echo 'PyInstaller failed to pack gProfiler code! See lines above. Make sure to check for SyntaxError as this is often the reason.'; exit 1; fi;
 
+# for aarch64 - build a patched version of staticx 0.13.6. we remove calls to getpwnam and getgrnam, for these end up doing dlopen()s which
+# crash the staticx bootloader. we don't need them anyway (all files in our staticx tar are uid 0 and we don't need the names translation)
+COPY scripts/staticx_patch.diff staticx_patch.diff
+RUN if [ $(uname -m) = "aarch64" ]; then \
+        git clone -b v0.13.6 https://github.com/JonathonReinhart/staticx.git && \
+        cd staticx && \
+        git reset --hard 819d8eafecbaab3646f70dfb1e3e19f6bbc017f8 && \
+        git apply ../staticx_patch.diff && \
+        python3 -m pip install . ; \
+    fi
+
 COPY ./scripts/list_needed_libs.sh ./scripts/list_needed_libs.sh
 # staticx packs dynamically linked app with all of their dependencies, it tries to figure out which dynamic libraries are need for its execution
 # in some cases, when the application is lazily loading some DSOs, staticx doesn't handle it.
 # we use list_needed_libs.sh to list the dynamic dependencies of *all* of our resources,
 # and make staticx pack them as well.
 # using scl here to get the proper LD_LIBRARY_PATH set
-# TODO: use staticx for aarch64 as well; currently it doesn't generate correct binaries when run over Docker emulation.
-RUN if [ $(uname -m) != "aarch64" ]; then source scl_source enable devtoolset-8 llvm-toolset-7 && libs=$(./scripts/list_needed_libs.sh) && staticx $libs dist/gprofiler dist/gprofiler; fi
+RUN if [ $(uname -m) != "aarch64" ]; then source scl_source enable devtoolset-8 llvm-toolset-7 ; fi; libs=$(./scripts/list_needed_libs.sh) && staticx $libs dist/gprofiler dist/gprofiler
 
 FROM scratch AS export-stage
 

scripts/list_needed_libs.sh

@@ -1,9 +1,10 @@
+
 #!/usr/bin/env bash
 #
 # Copyright (c) Granulate. All rights reserved.
 # Licensed under the AGPL3 License. See LICENSE.md in the project root for license information.
 #
-set -e
+set -euo pipefail
 
 # this file lists all dynamic dependenices of executables in gprofiler/resources.
 # we use it to let staticx know which libraries it should pack inside.
@@ -40,7 +41,11 @@ for f in $BINS ; do
     fi
 done
 
+printed=
 for l in $libs ; do
-    >&2 echo "found needed lib: $l"
-    echo -n " -l $l"
+    if ! echo "$printed" | grep -q "$l"; then
+        >&2 echo "found needed lib: $l"
+        echo -n " -l $l"
+        printed="$printed $l"
+    fi
 done

scripts/staticx_patch.diff

@@ -0,0 +1,28 @@
+diff --git a/libtar/decode.c b/libtar/decode.c
+index b34b1d9..d221c3a 100644
+--- a/libtar/decode.c
++++ b/libtar/decode.c
+@@ -44,11 +44,6 @@ uid_t
+ th_get_uid(const TAR *t)
+ {
+ 	int uid;
+-	struct passwd *pw;
+-
+-	pw = getpwnam(t->th_buf.uname);
+-	if (pw != NULL)
+-		return pw->pw_uid;
+ 
+ 	/* if the password entry doesn't exist */
+ 	sscanf(t->th_buf.uid, "%o", &uid);
+@@ -60,11 +55,6 @@ gid_t
+ th_get_gid(const TAR *t)
+ {
+ 	int gid;
+-	struct group *gr;
+-
+-	gr = getgrnam(t->th_buf.gname);
+-	if (gr != NULL)
+-		return gr->gr_gid;
+ 
+ 	/* if the group entry doesn't exist */
+ 	sscanf(t->th_buf.gid, "%o", &gid);