[PATCH 0/3] Forced ordering for DFS ELF dependency sorting (bug 28937)

[PATCH 0/3] Forced ordering for DFS ELF dependency sorting (bug 28937)

[Florian Weimer]

An Arch Linux test case showed that in some cases, dlclose could unload
mappings that are still in use.  Triggering this requires dependency
cycles, which is why I think this is somewhat rare.

Tested on i686-linux-gnu and x86_64-linux-gnu.

Thanks,
Florian

Florian Weimer (3):
  scripts/dso-ordering-test.py: Generate program run-time dependencies
  elf: Rename _dl_sort_maps parameter from skip to force_first
  elf: Implement force_first handling in _dl_sort_maps_dfs

 elf/dl-sort-maps.c           | 47 ++++++++++++++++++++++++------------
 elf/dso-sort-tests-1.def     |  7 ++++++
 scripts/dso-ordering-test.py | 14 +++++------
 sysdeps/generic/ldsodefs.h   |  6 +++--
 4 files changed, 50 insertions(+), 24 deletions(-)


base-commit: 453b88efe6fa79f5c7c6fccc3a520c75fdd43074
-- 
2.37.1

[PATCH 1/3] scripts/dso-ordering-test.py: Generate program run-time dependencies

[Florian Weimer]

The main program needs to depend on all shared objects, even objects
that have link-time dependencies among shared objects.  Filtering
out shared objects that already have an link-time dependencies is not
necessary here; make will do this automatically.
---
 scripts/dso-ordering-test.py | 14 +++++++-------
 1 file changed, 7 insertions(+), 7 deletions(-)

diff --git a/scripts/dso-ordering-test.py b/scripts/dso-ordering-test.py
index 2dd6bfda18..f6d22aa00e 100644
--- a/scripts/dso-ordering-test.py
+++ b/scripts/dso-ordering-test.py
@@ -707,13 +707,12 @@ def process_testcase(t):
                 "\t$(compile.c) $(OUTPUT_OPTION)\n")
         makefile.write (rule)
 
-        not_depended_objs = find_objs_not_depended_on(test_descr)
-        if not_depended_objs:
-            depstr = ""
-            for dep in not_depended_objs:
-                depstr += (" $(objpfx)" + test_subdir + "/"
-                           + test_name + "-" + dep + ".so")
-            makefile.write("$(objpfx)%s.out:%s\n" % (base_test_name, depstr))
+        # Ensure that all shared objects are built before running the
+        # test, whether there link-time dependencies or not.
+        depobjs = ["$(objpfx){}/{}-{}.so".format(test_subdir, test_name, dep)
+                   for dep in test_descr.objs]
+        makefile.write("$(objpfx){}.out: {}\n".format(
+            base_test_name, " ".join(depobjs)))
 
         # Add main executable to test-srcs
         makefile.write("test-srcs += %s/%s\n" % (test_subdir, test_name))
@@ -779,6 +778,7 @@ def process_testcase(t):
         # description string.
         for obj in test_descr.objs:
             src_name = test_name + "-" + obj + ".c"
+            print('Generating {}...'.format(src_name))
             f = open(testpfx_src + src_name, "w")
             if obj in test_descr.callrefs:
                 called_objs = test_descr.callrefs[obj]
-- 
2.37.1

Re: [PATCH 1/3] scripts/dso-ordering-test.py: Generate program run-time dependencies

[Florian Weimer]

* Florian Weimer via Libc-alpha:

> The main program needs to depend on all shared objects, even objects
> that have link-time dependencies among shared objects.  Filtering
> out shared objects that already have an link-time dependencies is not
> necessary here; make will do this automatically.
> ---
>  scripts/dso-ordering-test.py | 14 +++++++-------
>  1 file changed, 7 insertions(+), 7 deletions(-)
>
> diff --git a/scripts/dso-ordering-test.py b/scripts/dso-ordering-test.py
> index 2dd6bfda18..f6d22aa00e 100644
> --- a/scripts/dso-ordering-test.py
> +++ b/scripts/dso-ordering-test.py
> @@ -707,13 +707,12 @@ def process_testcase(t):
>                  "\t$(compile.c) $(OUTPUT_OPTION)\n")
>          makefile.write (rule)
>  
> -        not_depended_objs = find_objs_not_depended_on(test_descr)
> -        if not_depended_objs:
> -            depstr = ""
> -            for dep in not_depended_objs:
> -                depstr += (" $(objpfx)" + test_subdir + "/"
> -                           + test_name + "-" + dep + ".so")
> -            makefile.write("$(objpfx)%s.out:%s\n" % (base_test_name, depstr))
> +        # Ensure that all shared objects are built before running the
> +        # test, whether there link-time dependencies or not.
> +        depobjs = ["$(objpfx){}/{}-{}.so".format(test_subdir, test_name, dep)
> +                   for dep in test_descr.objs]
> +        makefile.write("$(objpfx){}.out: {}\n".format(
> +            base_test_name, " ".join(depobjs)))
>  
>          # Add main executable to test-srcs
>          makefile.write("test-srcs += %s/%s\n" % (test_subdir, test_name))
> @@ -779,6 +778,7 @@ def process_testcase(t):
>          # description string.
>          for obj in test_descr.objs:
>              src_name = test_name + "-" + obj + ".c"
> +            print('Generating {}...'.format(src_name))
>              f = open(testpfx_src + src_name, "w")
>              if obj in test_descr.callrefs:
>                  called_objs = test_descr.callrefs[obj]

Ugh, I'm going to remove that debugging leftover before committing.

Thanks,
Florian

Re: [PATCH 1/3] scripts/dso-ordering-test.py: Generate program run-time dependencies

[Adhemerval Zanella Netto]

On 15/08/22 11:30, Florian Weimer via Libc-alpha wrote:
> The main program needs to depend on all shared objects, even objects
> that have link-time dependencies among shared objects.  Filtering
> out shared objects that already have an link-time dependencies is not
> necessary here; make will do this automatically.

LGTM with the print debug leftover removed.

Reviewed-by: Adhemerval Zanella  <adhemerval.zanella@linaro.org>

> ---
>  scripts/dso-ordering-test.py | 14 +++++++-------
>  1 file changed, 7 insertions(+), 7 deletions(-)
> 
> diff --git a/scripts/dso-ordering-test.py b/scripts/dso-ordering-test.py
> index 2dd6bfda18..f6d22aa00e 100644
> --- a/scripts/dso-ordering-test.py
> +++ b/scripts/dso-ordering-test.py
> @@ -707,13 +707,12 @@ def process_testcase(t):
>                  "\t$(compile.c) $(OUTPUT_OPTION)\n")
>          makefile.write (rule)
>  
> -        not_depended_objs = find_objs_not_depended_on(test_descr)
> -        if not_depended_objs:
> -            depstr = ""
> -            for dep in not_depended_objs:
> -                depstr += (" $(objpfx)" + test_subdir + "/"
> -                           + test_name + "-" + dep + ".so")
> -            makefile.write("$(objpfx)%s.out:%s\n" % (base_test_name, depstr))
> +        # Ensure that all shared objects are built before running the
> +        # test, whether there link-time dependencies or not.
> +        depobjs = ["$(objpfx){}/{}-{}.so".format(test_subdir, test_name, dep)
> +                   for dep in test_descr.objs]
> +        makefile.write("$(objpfx){}.out: {}\n".format(
> +            base_test_name, " ".join(depobjs)))
>  
>          # Add main executable to test-srcs
>          makefile.write("test-srcs += %s/%s\n" % (test_subdir, test_name))
> @@ -779,6 +778,7 @@ def process_testcase(t):
>          # description string.
>          for obj in test_descr.objs:
>              src_name = test_name + "-" + obj + ".c"
> +            print('Generating {}...'.format(src_name))
>              f = open(testpfx_src + src_name, "w")
>              if obj in test_descr.callrefs:
>                  called_objs = test_descr.callrefs[obj]

[PATCH 2/3] elf: Rename _dl_sort_maps parameter from skip to force_first

[Florian Weimer]

The new implementation will not be able to skip an arbitrary number
of objects.
---
 elf/dl-sort-maps.c         | 14 +++++++-------
 sysdeps/generic/ldsodefs.h |  6 ++++--
 2 files changed, 11 insertions(+), 9 deletions(-)

diff --git a/elf/dl-sort-maps.c b/elf/dl-sort-maps.c
index 96638d7ed1..5b550b1e94 100644
--- a/elf/dl-sort-maps.c
+++ b/elf/dl-sort-maps.c
@@ -27,12 +27,12 @@
    If FOR_FINI is true, this is called for finishing an object.  */
 static void
 _dl_sort_maps_original (struct link_map **maps, unsigned int nmaps,
-			unsigned int skip, bool for_fini)
+			bool force_first, bool for_fini)
 {
   /* Allows caller to do the common optimization of skipping the first map,
      usually the main binary.  */
-  maps += skip;
-  nmaps -= skip;
+  maps += force_first;
+  nmaps -= force_first;
 
   /* A list of one element need not be sorted.  */
   if (nmaps <= 1)
@@ -182,7 +182,7 @@ dfs_traversal (struct link_map ***rpo, struct link_map *map,
 
 static void
 _dl_sort_maps_dfs (struct link_map **maps, unsigned int nmaps,
-		   unsigned int skip __attribute__ ((unused)), bool for_fini)
+		   bool force_first __attribute__ ((unused)), bool for_fini)
 {
   for (int i = nmaps - 1; i >= 0; i--)
     maps[i]->l_visited = 0;
@@ -286,7 +286,7 @@ _dl_sort_maps_init (void)
 
 void
 _dl_sort_maps (struct link_map **maps, unsigned int nmaps,
-	       unsigned int skip, bool for_fini)
+	       bool force_first, bool for_fini)
 {
   /* It can be tempting to use a static function pointer to store and call
      the current selected sorting algorithm routine, but experimentation
@@ -296,9 +296,9 @@ _dl_sort_maps (struct link_map **maps, unsigned int nmaps,
      input cases. A simple if-case with direct function calls appears to
      be the fastest.  */
   if (__glibc_likely (GLRO(dl_dso_sort_algo) == dso_sort_algorithm_original))
-    _dl_sort_maps_original (maps, nmaps, skip, for_fini);
+    _dl_sort_maps_original (maps, nmaps, force_first, for_fini);
   else
-    _dl_sort_maps_dfs (maps, nmaps, skip, for_fini);
+    _dl_sort_maps_dfs (maps, nmaps, force_first, for_fini);
 }
 
 #endif /* HAVE_TUNABLES.  */
diff --git a/sysdeps/generic/ldsodefs.h b/sysdeps/generic/ldsodefs.h
index 050a3032de..6b256b8388 100644
--- a/sysdeps/generic/ldsodefs.h
+++ b/sysdeps/generic/ldsodefs.h
@@ -1048,9 +1048,11 @@ extern void _dl_init (struct link_map *main_map, int argc, char **argv,
    initializer functions have completed.  */
 extern void _dl_fini (void) attribute_hidden;
 
-/* Sort array MAPS according to dependencies of the contained objects.  */
+/* Sort array MAPS according to dependencies of the contained objects.
+   If FORCE_FIRST, MAPS[0] keeps its place even if the dependencies
+   say otherwise.  */
 extern void _dl_sort_maps (struct link_map **maps, unsigned int nmaps,
-			   unsigned int skip, bool for_fini) attribute_hidden;
+			   bool force_first, bool for_fini) attribute_hidden;
 
 /* The dynamic linker calls this function before and having changing
    any shared object mappings.  The `r_state' member of `struct r_debug'
-- 
2.37.1

Re: [PATCH 2/3] elf: Rename _dl_sort_maps parameter from skip to force_first

[Adhemerval Zanella Netto]

On 15/08/22 11:30, Florian Weimer via Libc-alpha wrote:
> The new implementation will not be able to skip an arbitrary number
> of objects.

LGTM, thanks.

Reviewed-by: Adhemerval Zanella  <adhemerval.zanella@linaro.org>

> ---
>  elf/dl-sort-maps.c         | 14 +++++++-------
>  sysdeps/generic/ldsodefs.h |  6 ++++--
>  2 files changed, 11 insertions(+), 9 deletions(-)
> 
> diff --git a/elf/dl-sort-maps.c b/elf/dl-sort-maps.c
> index 96638d7ed1..5b550b1e94 100644
> --- a/elf/dl-sort-maps.c
> +++ b/elf/dl-sort-maps.c
> @@ -27,12 +27,12 @@
>     If FOR_FINI is true, this is called for finishing an object.  */
>  static void
>  _dl_sort_maps_original (struct link_map **maps, unsigned int nmaps,
> -			unsigned int skip, bool for_fini)
> +			bool force_first, bool for_fini)
>  {
>    /* Allows caller to do the common optimization of skipping the first map,
>       usually the main binary.  */
> -  maps += skip;
> -  nmaps -= skip;
> +  maps += force_first;
> +  nmaps -= force_first;
>  
>    /* A list of one element need not be sorted.  */
>    if (nmaps <= 1)
> @@ -182,7 +182,7 @@ dfs_traversal (struct link_map ***rpo, struct link_map *map,
>  
>  static void
>  _dl_sort_maps_dfs (struct link_map **maps, unsigned int nmaps,
> -		   unsigned int skip __attribute__ ((unused)), bool for_fini)
> +		   bool force_first __attribute__ ((unused)), bool for_fini)
>  {
>    for (int i = nmaps - 1; i >= 0; i--)
>      maps[i]->l_visited = 0;
> @@ -286,7 +286,7 @@ _dl_sort_maps_init (void)
>  
>  void
>  _dl_sort_maps (struct link_map **maps, unsigned int nmaps,
> -	       unsigned int skip, bool for_fini)
> +	       bool force_first, bool for_fini)
>  {
>    /* It can be tempting to use a static function pointer to store and call
>       the current selected sorting algorithm routine, but experimentation
> @@ -296,9 +296,9 @@ _dl_sort_maps (struct link_map **maps, unsigned int nmaps,
>       input cases. A simple if-case with direct function calls appears to
>       be the fastest.  */
>    if (__glibc_likely (GLRO(dl_dso_sort_algo) == dso_sort_algorithm_original))
> -    _dl_sort_maps_original (maps, nmaps, skip, for_fini);
> +    _dl_sort_maps_original (maps, nmaps, force_first, for_fini);
>    else
> -    _dl_sort_maps_dfs (maps, nmaps, skip, for_fini);
> +    _dl_sort_maps_dfs (maps, nmaps, force_first, for_fini);
>  }
>  
>  #endif /* HAVE_TUNABLES.  */
> diff --git a/sysdeps/generic/ldsodefs.h b/sysdeps/generic/ldsodefs.h
> index 050a3032de..6b256b8388 100644
> --- a/sysdeps/generic/ldsodefs.h
> +++ b/sysdeps/generic/ldsodefs.h
> @@ -1048,9 +1048,11 @@ extern void _dl_init (struct link_map *main_map, int argc, char **argv,
>     initializer functions have completed.  */
>  extern void _dl_fini (void) attribute_hidden;
>  
> -/* Sort array MAPS according to dependencies of the contained objects.  */
> +/* Sort array MAPS according to dependencies of the contained objects.
> +   If FORCE_FIRST, MAPS[0] keeps its place even if the dependencies
> +   say otherwise.  */
>  extern void _dl_sort_maps (struct link_map **maps, unsigned int nmaps,
> -			   unsigned int skip, bool for_fini) attribute_hidden;
> +			   bool force_first, bool for_fini) attribute_hidden;
>  
>  /* The dynamic linker calls this function before and having changing
>     any shared object mappings.  The `r_state' member of `struct r_debug'

[PATCH 3/3] elf: Implement force_first handling in _dl_sort_maps_dfs

[Florian Weimer]

As documented in a comment _dl_close_worker, the skipping is actually
needed for correctness.  It also seems less surprising if the
just-opened object is always initialized last, even in the presence
of cycles.
---
 elf/dl-sort-maps.c       | 35 ++++++++++++++++++++++++++---------
 elf/dso-sort-tests-1.def |  7 +++++++
 2 files changed, 33 insertions(+), 9 deletions(-)

diff --git a/elf/dl-sort-maps.c b/elf/dl-sort-maps.c
index 5b550b1e94..cd2d9c93fc 100644
--- a/elf/dl-sort-maps.c
+++ b/elf/dl-sort-maps.c
@@ -182,8 +182,9 @@ dfs_traversal (struct link_map ***rpo, struct link_map *map,
 
 static void
 _dl_sort_maps_dfs (struct link_map **maps, unsigned int nmaps,
-		   bool force_first __attribute__ ((unused)), bool for_fini)
+		   bool force_first, bool for_fini)
 {
+  struct link_map *first_map = maps[0];
   for (int i = nmaps - 1; i >= 0; i--)
     maps[i]->l_visited = 0;
 
@@ -208,14 +209,6 @@ _dl_sort_maps_dfs (struct link_map **maps, unsigned int nmaps,
      Adjusting the order so that maps[0] is last traversed naturally avoids
      this problem.
 
-     Further, the old "optimization" of skipping the main object at maps[0]
-     from the call-site (i.e. _dl_sort_maps(maps+1,nmaps-1)) is in general
-     no longer valid, since traversing along object dependency-links
-     may "find" the main object even when it is not included in the initial
-     order (e.g. a dlopen()'ed shared object can have circular dependencies
-     linked back to itself). In such a case, traversing N-1 objects will
-     create a N-object result, and raise problems.
-
      To summarize, just passing in the full list, and iterating from back
      to front makes things much more straightforward.  */
 
@@ -274,6 +267,30 @@ _dl_sort_maps_dfs (struct link_map **maps, unsigned int nmaps,
     }
 
   memcpy (maps, rpo, sizeof (struct link_map *) * nmaps);
+
+  /* Skipping the first object at maps[0] is not valid in general,
+     since traversing along object dependency-links may "find" that
+     first object even when it is not included in the initial order
+     (e.g. a dlopen()'ed shared object can have circular dependencies
+     linked back to itself). In such a case, traversing N-1 objects
+     will create a N-object result, and raise problems.  Instead,
+     force the object back into first place after sorting.  */
+  if (force_first && maps[0] != first_map)
+    {
+      struct link_map *saved = maps[0];
+      maps[0] = first_map;
+      int i = 1;
+      while (true)
+	{
+	  assert (i < nmaps);
+	  struct link_map *current = maps[i];
+	  maps[i] = saved;
+	  if (current == first_map)
+	     break;
+	  saved = current;
+	  ++i;
+	}
+    }
 }
 
 void
diff --git a/elf/dso-sort-tests-1.def b/elf/dso-sort-tests-1.def
index 5f7f18ef27..4bf9052db1 100644
--- a/elf/dso-sort-tests-1.def
+++ b/elf/dso-sort-tests-1.def
@@ -64,3 +64,10 @@ output: b>a>{}<a<b
 tst-bz15311: {+a;+e;+f;+g;+d;%d;-d;-g;-f;-e;-a};a->b->c->d;d=>[ba];c=>a;b=>e=>a;c=>f=>b;d=>g=>c
 output(glibc.rtld.dynamic_sort=1): {+a[d>c>b>a>];+e[e>];+f[f>];+g[g>];+d[];%d(b(e(a()))a()g(c(a()f(b(e(a()))))));-d[];-g[];-f[];-e[];-a[<a<c<d<g<f<b<e];}
 output(glibc.rtld.dynamic_sort=2): {+a[d>c>b>a>];+e[e>];+f[f>];+g[g>];+d[];%d(b(e(a()))a()g(c(a()f(b(e(a()))))));-d[];-g[];-f[];-e[];-a[<g<f<a<b<c<d<e];}
+
+# Test that even in the presence of dependency loops involving dlopen'ed
+# object, that object is initialized last (and not unloaded prematurely).
+# Final destructor order is indeterminate due to the cycle.
+tst-bz28937: {+a;+b;-b;+c;%c};a->a1;a->a2;a2->a;b->b1;c->a1;c=>a1
+output(glibc.rtld.dynamic_sort=1): {+a[a2>a1>a>];+b[b1>b>];-b[<b<b1];+c[c>];%c(a1());}<a<a2<c<a1
+output(glibc.rtld.dynamic_sort=2): {+a[a2>a1>a>];+b[b1>b>];-b[<b<b1];+c[c>];%c(a1());}<a2<a<c<a1
-- 
2.37.1

Re: [PATCH 3/3] elf: Implement force_first handling in _dl_sort_maps_dfs

[Adhemerval Zanella Netto]

On 15/08/22 11:30, Florian Weimer via Libc-alpha wrote:
> As documented in a comment _dl_close_worker, the skipping is actually
> needed for correctness.  It also seems less surprising if the
> just-opened object is always initialized last, even in the presence
> of cycles.

I think it is BZ#28937, isn't? Also could you extend the explanation as you
did in the last comment, the initial phrase sounds confusing.  Maybe extend
the comment to say that not _dl_sort_maps_dfs will move the main object to
front, so where previous you have the maps input as:

  maps[0].l_name=elf/tst-bz28937-dir/tst-bz28937-a.so
  maps[1].l_name=[...]/elf/tst-bz28937-dir/tst-bz28937-a1.so
  maps[2].l_name=[...]/elf/tst-bz28937-dir/tst-bz28937-a2.so
  maps[3].l_name=./libc.so.6
  maps[4].l_name=elf/ld.so

It will not be properly sorted as:

  maps[0].l_name=elf/tst-bz28937-dir/tst-bz28937-a.so
  maps[1].l_name=[...]/elf/tst-bz28937-dir/tst-bz28937-a1.so
  maps[2].l_name=[...]/elf/tst-bz28937-dir/tst-bz28937-a2.so
  maps[3].l_name=./libc.so.6
  maps[4].l_name=elf/ld.so

Instead of wrongly:

  maps[0].l_name=[...]/elf/tst-bz28937-dir/tst-bz28937-a1.so
  maps[1].l_name=[...]/elf/tst-bz28937-dir/tst-bz28937-a2.so
  maps[2].l_name=elf/tst-bz28937-dir/tst-bz28937-a.so
  maps[3].l_name=./libc.so.6
  maps[4].l_name=elf/ld.so

> ---
>  elf/dl-sort-maps.c       | 35 ++++++++++++++++++++++++++---------
>  elf/dso-sort-tests-1.def |  7 +++++++
>  2 files changed, 33 insertions(+), 9 deletions(-)
> 
> diff --git a/elf/dl-sort-maps.c b/elf/dl-sort-maps.c
> index 5b550b1e94..cd2d9c93fc 100644
> --- a/elf/dl-sort-maps.c
> +++ b/elf/dl-sort-maps.c
> @@ -182,8 +182,9 @@ dfs_traversal (struct link_map ***rpo, struct link_map *map,
>  
>  static void
>  _dl_sort_maps_dfs (struct link_map **maps, unsigned int nmaps,
> -		   bool force_first __attribute__ ((unused)), bool for_fini)
> +		   bool force_first, bool for_fini)
>  {
> +  struct link_map *first_map = maps[0];

Move this to where it is actually used.

>    for (int i = nmaps - 1; i >= 0; i--)
>      maps[i]->l_visited = 0;
>  
> @@ -208,14 +209,6 @@ _dl_sort_maps_dfs (struct link_map **maps, unsigned int nmaps,
>       Adjusting the order so that maps[0] is last traversed naturally avoids
>       this problem.
>  
> -     Further, the old "optimization" of skipping the main object at maps[0]
> -     from the call-site (i.e. _dl_sort_maps(maps+1,nmaps-1)) is in general
> -     no longer valid, since traversing along object dependency-links
> -     may "find" the main object even when it is not included in the initial
> -     order (e.g. a dlopen()'ed shared object can have circular dependencies> -     linked back to itself). In such a case, traversing N-1 objects will
> -     create a N-object result, and raise problems.
> -
>       To summarize, just passing in the full list, and iterating from back
>       to front makes things much more straightforward.  */
>  
> @@ -274,6 +267,30 @@ _dl_sort_maps_dfs (struct link_map **maps, unsigned int nmaps,
>      }
>  
>    memcpy (maps, rpo, sizeof (struct link_map *) * nmaps);
> +
> +  /* Skipping the first object at maps[0] is not valid in general,
> +     since traversing along object dependency-links may "find" that
> +     first object even when it is not included in the initial order
> +     (e.g. a dlopen()'ed shared object can have circular dependencies
> +     linked back to itself). In such a case, traversing N-1 objects

Double space after period and think we do not reference symbol using '()'.

> +     will create a N-object result, and raise problems.  Instead,
> +     force the object back into first place after sorting.  */
> +  if (force_first && maps[0] != first_map)
> +    {
> +      struct link_map *saved = maps[0];
> +      maps[0] = first_map;
> +      int i = 1;
> +      while (true)
> +	{
> +	  assert (i < nmaps);
> +	  struct link_map *current = maps[i];
> +	  maps[i] = saved;
> +	  if (current == first_map)
> +	     break;
> +	  saved = current;
> +	  ++i;
> +	}
> +    }
>  }

It sounds reasonable to keep the main object as the initial map, although
it would slow down a bit normal dlclose.  I think it would be possible
to optimize the memory move with memmove here, although not sure if it is
worth.

>  
>  void
> diff --git a/elf/dso-sort-tests-1.def b/elf/dso-sort-tests-1.def
> index 5f7f18ef27..4bf9052db1 100644
> --- a/elf/dso-sort-tests-1.def
> +++ b/elf/dso-sort-tests-1.def
> @@ -64,3 +64,10 @@ output: b>a>{}<a<b
>  tst-bz15311: {+a;+e;+f;+g;+d;%d;-d;-g;-f;-e;-a};a->b->c->d;d=>[ba];c=>a;b=>e=>a;c=>f=>b;d=>g=>c
>  output(glibc.rtld.dynamic_sort=1): {+a[d>c>b>a>];+e[e>];+f[f>];+g[g>];+d[];%d(b(e(a()))a()g(c(a()f(b(e(a()))))));-d[];-g[];-f[];-e[];-a[<a<c<d<g<f<b<e];}
>  output(glibc.rtld.dynamic_sort=2): {+a[d>c>b>a>];+e[e>];+f[f>];+g[g>];+d[];%d(b(e(a()))a()g(c(a()f(b(e(a()))))));-d[];-g[];-f[];-e[];-a[<g<f<a<b<c<d<e];}
> +
> +# Test that even in the presence of dependency loops involving dlopen'ed
> +# object, that object is initialized last (and not unloaded prematurely).
> +# Final destructor order is indeterminate due to the cycle.
> +tst-bz28937: {+a;+b;-b;+c;%c};a->a1;a->a2;a2->a;b->b1;c->a1;c=>a1

So main program:

  1. dlopen 'a' and 'b';
  2. dclose 'b';
  3. dlopen 'c';
  4. dlsym 'c' and calls fn_a from 'c';

And we have a circle dependency where a depends of a2 and a2 depends on 'a'.

Do we need to add a test for multiple circles? For instance where you have
either another disjointed circle ({+d};d->d2;d2->d) and/or another circle
in same dependency chain (a1->b;b1->a)?

> +output(glibc.rtld.dynamic_sort=1): {+a[a2>a1>a>];+b[b1>b>];-b[<b<b1];+c[c>];%c(a1());}<a<a2<c<a1
> +output(glibc.rtld.dynamic_sort=2): {+a[a2>a1>a>];+b[b1>b>];-b[<b<b1];+c[c>];%c(a1());}<a2<a<c<a1

Re: [PATCH 3/3] elf: Implement force_first handling in _dl_sort_maps_dfs

[Adhemerval Zanella Netto]

On 31/08/22 13:37, Adhemerval Zanella Netto wrote:
> 
> 
> On 15/08/22 11:30, Florian Weimer via Libc-alpha wrote:
>> As documented in a comment _dl_close_worker, the skipping is actually
>> needed for correctness.  It also seems less surprising if the
>> just-opened object is always initialized last, even in the presence
>> of cycles.
> 
> I think it is BZ#28937, isn't? Also could you extend the explanation as you
> did in the last comment, the initial phrase sounds confusing.  Maybe extend
> the comment to say that not _dl_sort_maps_dfs will move the main object to
> front, so where previous you have the maps input as:
> 
>   maps[0].l_name=elf/tst-bz28937-dir/tst-bz28937-a.so
>   maps[1].l_name=[...]/elf/tst-bz28937-dir/tst-bz28937-a1.so
>   maps[2].l_name=[...]/elf/tst-bz28937-dir/tst-bz28937-a2.so
>   maps[3].l_name=./libc.so.6
>   maps[4].l_name=elf/ld.so
> 
> It will not be properly sorted as:

It will *now* be ...

Re: [PATCH 3/3] elf: Implement force_first handling in _dl_sort_maps_dfs

[Florian Weimer]

* Adhemerval Zanella Netto:

> On 15/08/22 11:30, Florian Weimer via Libc-alpha wrote:
>> As documented in a comment _dl_close_worker, the skipping is actually
>> needed for correctness.  It also seems less surprising if the
>> just-opened object is always initialized last, even in the presence
>> of cycles.
>
> I think it is BZ#28937, isn't? Also could you extend the explanation as you
> did in the last comment, the initial phrase sounds confusing.

I have expanded the commit message.

> Maybe extend the comment to say that not _dl_sort_maps_dfs will move
> the main object to front, so where previous you have the maps input
> as:
>
>   maps[0].l_name=elf/tst-bz28937-dir/tst-bz28937-a.so
>   maps[1].l_name=[...]/elf/tst-bz28937-dir/tst-bz28937-a1.so
>   maps[2].l_name=[...]/elf/tst-bz28937-dir/tst-bz28937-a2.so
>   maps[3].l_name=./libc.so.6
>   maps[4].l_name=elf/ld.so
>
> It will not be properly sorted as:
>
>   maps[0].l_name=elf/tst-bz28937-dir/tst-bz28937-a.so
>   maps[1].l_name=[...]/elf/tst-bz28937-dir/tst-bz28937-a1.so
>   maps[2].l_name=[...]/elf/tst-bz28937-dir/tst-bz28937-a2.so
>   maps[3].l_name=./libc.so.6
>   maps[4].l_name=elf/ld.so
>
> Instead of wrongly:
>
>   maps[0].l_name=[...]/elf/tst-bz28937-dir/tst-bz28937-a1.so
>   maps[1].l_name=[...]/elf/tst-bz28937-dir/tst-bz28937-a2.so
>   maps[2].l_name=elf/tst-bz28937-dir/tst-bz28937-a.so
>   maps[3].l_name=./libc.so.6
>   maps[4].l_name=elf/ld.so

I think with just three elements, it's a bit misleading because a cycle
of this size is rotated correctly by the naive approach.

>> diff --git a/elf/dl-sort-maps.c b/elf/dl-sort-maps.c
>> index 5b550b1e94..cd2d9c93fc 100644
>> --- a/elf/dl-sort-maps.c
>> +++ b/elf/dl-sort-maps.c
>> @@ -182,8 +182,9 @@ dfs_traversal (struct link_map ***rpo, struct link_map *map,
>>  
>>  static void
>>  _dl_sort_maps_dfs (struct link_map **maps, unsigned int nmaps,
>> -		   bool force_first __attribute__ ((unused)), bool for_fini)
>> +		   bool force_first, bool for_fini)
>>  {
>> +  struct link_map *first_map = maps[0];
>
> Move this to where it is actually used.

We need to copy it before it's overwritten by the sort.

>> +  /* Skipping the first object at maps[0] is not valid in general,
>> +     since traversing along object dependency-links may "find" that
>> +     first object even when it is not included in the initial order
>> +     (e.g. a dlopen()'ed shared object can have circular dependencies
>> +     linked back to itself). In such a case, traversing N-1 objects
>
> Double space after period and think we do not reference symbol using '()'.

Fixed (although I just copied that part of the comment).

>> +     will create a N-object result, and raise problems.  Instead,
>> +     force the object back into first place after sorting.  */
>> +  if (force_first && maps[0] != first_map)
>> +    {
>> +      struct link_map *saved = maps[0];
>> +      maps[0] = first_map;
>> +      int i = 1;
>> +      while (true)
>> +	{
>> +	  assert (i < nmaps);
>> +	  struct link_map *current = maps[i];
>> +	  maps[i] = saved;
>> +	  if (current == first_map)
>> +	     break;
>> +	  saved = current;
>> +	  ++i;
>> +	}
>> +    }
>>  }
>
> It sounds reasonable to keep the main object as the initial map, although
> it would slow down a bit normal dlclose.  I think it would be possible
> to optimize the memory move with memmove here, although not sure if it is
> worth.

If we make the assert less precise, memmove actually simplifies the
code.  I've made the change.

>>  void
>> diff --git a/elf/dso-sort-tests-1.def b/elf/dso-sort-tests-1.def
>> index 5f7f18ef27..4bf9052db1 100644
>> --- a/elf/dso-sort-tests-1.def
>> +++ b/elf/dso-sort-tests-1.def
>> @@ -64,3 +64,10 @@ output: b>a>{}<a<b
>>  tst-bz15311: {+a;+e;+f;+g;+d;%d;-d;-g;-f;-e;-a};a->b->c->d;d=>[ba];c=>a;b=>e=>a;c=>f=>b;d=>g=>c
>>  output(glibc.rtld.dynamic_sort=1): {+a[d>c>b>a>];+e[e>];+f[f>];+g[g>];+d[];%d(b(e(a()))a()g(c(a()f(b(e(a()))))));-d[];-g[];-f[];-e[];-a[<a<c<d<g<f<b<e];}
>>  output(glibc.rtld.dynamic_sort=2): {+a[d>c>b>a>];+e[e>];+f[f>];+g[g>];+d[];%d(b(e(a()))a()g(c(a()f(b(e(a()))))));-d[];-g[];-f[];-e[];-a[<g<f<a<b<c<d<e];}
>> +
>> +# Test that even in the presence of dependency loops involving dlopen'ed
>> +# object, that object is initialized last (and not unloaded prematurely).
>> +# Final destructor order is indeterminate due to the cycle.
>> +tst-bz28937: {+a;+b;-b;+c;%c};a->a1;a->a2;a2->a;b->b1;c->a1;c=>a1
>
> So main program:
>
>   1. dlopen 'a' and 'b';
>   2. dclose 'b';
>   3. dlopen 'c';
>   4. dlsym 'c' and calls fn_a from 'c';
>
> And we have a circle dependency where a depends of a2 and a2 depends on 'a'.
>
> Do we need to add a test for multiple circles? For instance where you have
> either another disjointed circle ({+d};d->d2;d2->d) and/or another circle
> in same dependency chain (a1->b;b1->a)?

Could you add this as a follow-up patch?  It does not seem strictly
related (and I think we already have other tests for unloading cycles).

Thanks,
Florian