[PATCH Fix incorrect double-checked locking related to _res_hconf.initialized. [BZ #20477]

[PATCH Fix incorrect double-checked locking related to _res_hconf.initialized. [BZ #20477]

[Torvald Riegel]

[-- Attachment #1: Type: text/plain, Size: 874 bytes --]

_res_hconf.initialized was not suitable for use in a multi-threaded
environment due to the lack of atomics and memory barriers.  Use of it
was also unnecessary because _res_hconf_init did the right thing by
using __libc_once.  This patch fixes the glibc-internal uses by just
calling _res_hconf_init unconditionally, and switches to a release MO
atomic store for _res_hconf.initialized to fix the glibc side of the
synchronization problem (which will maintain backward compatibility, but
cannot fix the lack of acquire MO on any glibc-external loads).
    
        [BZ #20477]
        * resolv/res_hconf.c (do_init): Use atomic access.
        * resolv/res_hconf.h: Add comments.
        * nscd/aicache.c (addhstaiX): Call _res_hconf_init unconditionally.
        * nss/getXXbyYY_r.c (REENTRANT_NAME): Likewise.
        * sysdeps/posix/getaddrinfo.c (gaih_inet): Likewise.


[-- Attachment #2: res_hconf.patch --]
[-- Type: text/x-patch, Size: 3710 bytes --]

commit 44d3ca295ade8cb11b1473a7009705f6ec99d9bf
Author: Torvald Riegel <triegel@redhat.com>
Date:   Wed Aug 17 13:56:11 2016 +0200

    Fix incorrect double-checked locking related to _res_hconf.initialized.
    
    _res_hconf.initialized was not suitable for use in a multi-threaded
    environment due to the lack of atomics and memory barriers.  Use of it was
    also unnecessary because _res_hconf_init did the right thing by using
    __libc_once.  This patch fixes the glibc-internal uses by just calling
    _res_hconf_init unconditionally, and switches to a release MO atomic store
    for _res_hconf.initialized to fix the glibc side of the synchronization
    problem (which will maintain backward compatibility, but cannot fix the
    lack of acquire MO on any glibc-external loads).
    
    	[BZ #20477]
    	* resolv/res_hconf.c (do_init): Use atomic access.
    	* resolv/res_hconf.h: Add comments.
    	* nscd/aicache.c (addhstaiX): Call _res_hconf_init unconditionally.
    	* nss/getXXbyYY_r.c (REENTRANT_NAME): Likewise.
    	* sysdeps/posix/getaddrinfo.c (gaih_inet): Likewise.

diff --git a/nscd/aicache.c b/nscd/aicache.c
index a2e6cf8..32c8f57 100644
--- a/nscd/aicache.c
+++ b/nscd/aicache.c
@@ -101,8 +101,7 @@ addhstaiX (struct database_dyn *db, int fd, request_header *req,
   nip = hosts_database;
 
   /* Initialize configurations.  */
-  if (__glibc_unlikely (!_res_hconf.initialized))
-    _res_hconf_init ();
+  _res_hconf_init ();
   if (__res_maybe_init (&_res, 0) == -1)
     no_more = 1;
 
diff --git a/nss/getXXbyYY_r.c b/nss/getXXbyYY_r.c
index 93af253..18d3ad6 100644
--- a/nss/getXXbyYY_r.c
+++ b/nss/getXXbyYY_r.c
@@ -274,8 +274,7 @@ INTERNAL (REENTRANT_NAME) (ADD_PARAMS, LOOKUP_TYPE *resbuf, char *buffer,
 	    }
 #endif /* need _res */
 #ifdef NEED__RES_HCONF
-	  if (!_res_hconf.initialized)
-	    _res_hconf_init ();
+	  _res_hconf_init ();
 #endif /* need _res_hconf */
 
 	  void *tmp_ptr = fct.l;
diff --git a/resolv/res_hconf.c b/resolv/res_hconf.c
index 5cd1289..093c268 100644
--- a/resolv/res_hconf.c
+++ b/resolv/res_hconf.c
@@ -348,7 +348,8 @@ do_init (void)
       arg_trimdomain_list (ENV_TRIM_OVERR, 1, envval);
     }
 
-  _res_hconf.initialized = 1;
+  /* See comments on the declaration of _res_hconf.  */
+  atomic_store_release (&_res_hconf.initialized, 1);
 }
 
 
diff --git a/resolv/res_hconf.h b/resolv/res_hconf.h
index b97734d..a3d23f3 100644
--- a/resolv/res_hconf.h
+++ b/resolv/res_hconf.h
@@ -25,6 +25,15 @@
 
 struct hconf
 {
+  /* We keep the INITIALIZED member only for backwards compatibility.  New
+     code should just call _res_hconf_init unconditionally.  For this field
+     to be used safely, users must ensure that either (1) a call to
+     _res_hconf_init happens-before any load from INITIALIZED, or (2) an
+     assignment of zero to INITIALIZED happens-before any load from it, and
+     these loads use acquire MO if the intent is to skip calling
+     _res_hconf_init if the load returns a nonzero value.  Such acquire MO
+     loads will then synchronize with the release MO store to INITIALIZED
+     in do_init in res_hconf.c; see pthread_once for more detail.  */
   int initialized;
   int unused1;
   int unused2[4];
diff --git a/sysdeps/posix/getaddrinfo.c b/sysdeps/posix/getaddrinfo.c
index 574ce08..09fbc83 100644
--- a/sysdeps/posix/getaddrinfo.c
+++ b/sysdeps/posix/getaddrinfo.c
@@ -816,8 +816,7 @@ gaih_inet (const char *name, const struct gaih_service *service,
 	  nip = __nss_hosts_database;
 
 	  /* Initialize configurations.  */
-	  if (__glibc_unlikely (!_res_hconf.initialized))
-	    _res_hconf_init ();
+	  _res_hconf_init ();
 	  if (__res_maybe_init (&_res, 0) == -1)
 	    no_more = 1;
 

Re: [PATCH Fix incorrect double-checked locking related to _res_hconf.initialized. [BZ #20477]

[Florian Weimer]

On 08/17/2016 02:15 PM, Torvald Riegel wrote:
> _res_hconf.initialized was not suitable for use in a multi-threaded
> environment due to the lack of atomics and memory barriers.  Use of it
> was also unnecessary because _res_hconf_init did the right thing by
> using __libc_once.  This patch fixes the glibc-internal uses by just
> calling _res_hconf_init unconditionally, and switches to a release MO
> atomic store for _res_hconf.initialized to fix the glibc side of the
> synchronization problem (which will maintain backward compatibility, but
> cannot fix the lack of acquire MO on any glibc-external loads).
>
>         [BZ #20477]
>         * resolv/res_hconf.c (do_init): Use atomic access.
>         * resolv/res_hconf.h: Add comments.
>         * nscd/aicache.c (addhstaiX): Call _res_hconf_init unconditionally.
>         * nss/getXXbyYY_r.c (REENTRANT_NAME): Likewise.
>         * sysdeps/posix/getaddrinfo.c (gaih_inet): Likewise.

Looks good to me.  Please commit.

Thanks,
Florian

Re: [PATCH Fix incorrect double-checked locking related to _res_hconf.initialized. [BZ #20477]

[Andreas Schwab]

On Aug 17 2016, Torvald Riegel <triegel@redhat.com> wrote:

> diff --git a/resolv/res_hconf.h b/resolv/res_hconf.h
> index b97734d..a3d23f3 100644
> --- a/resolv/res_hconf.h
> +++ b/resolv/res_hconf.h
> @@ -25,6 +25,15 @@
>  
>  struct hconf
>  {
> +  /* We keep the INITIALIZED member only for backwards compatibility.  New
> +     code should just call _res_hconf_init unconditionally.  For this field
> +     to be used safely, users must ensure that either (1) a call to
> +     _res_hconf_init happens-before any load from INITIALIZED, or (2) an
                               ^
> +     assignment of zero to INITIALIZED happens-before any load from it, and
                                                 ^

Those hyphens look wrong.

Andreas.

-- 
Andreas Schwab, SUSE Labs, schwab@suse.de
GPG Key fingerprint = 0196 BAD8 1CE9 1970 F4BE  1748 E4D4 88E3 0EEA B9D7
"And now for something completely different."

Re: [PATCH Fix incorrect double-checked locking related to _res_hconf.initialized. [BZ #20477]

[Florian Weimer]

On 08/17/2016 04:51 PM, Andreas Schwab wrote:
> On Aug 17 2016, Torvald Riegel <triegel@redhat.com> wrote:
>
>> diff --git a/resolv/res_hconf.h b/resolv/res_hconf.h
>> index b97734d..a3d23f3 100644
>> --- a/resolv/res_hconf.h
>> +++ b/resolv/res_hconf.h
>> @@ -25,6 +25,15 @@
>>
>>  struct hconf
>>  {
>> +  /* We keep the INITIALIZED member only for backwards compatibility.  New
>> +     code should just call _res_hconf_init unconditionally.  For this field
>> +     to be used safely, users must ensure that either (1) a call to
>> +     _res_hconf_init happens-before any load from INITIALIZED, or (2) an
>                                ^
>> +     assignment of zero to INITIALIZED happens-before any load from it, and
>                                                  ^
>
> Those hyphens look wrong.

“happens-before” is third-person singular of the verb “happen-before”. 
It's a technical term widely used while discussing memory models. 
Here's an example:

   <http://docs.oracle.com/javase/specs/jls/se7/html/jls-17.html#jls-17.4.5>

Florian

Re: [PATCH Fix incorrect double-checked locking related to _res_hconf.initialized. [BZ #20477]

[Andreas Schwab]

On Aug 17 2016, Florian Weimer <fweimer@redhat.com> wrote:

> On 08/17/2016 04:51 PM, Andreas Schwab wrote:
>> On Aug 17 2016, Torvald Riegel <triegel@redhat.com> wrote:
>>
>>> diff --git a/resolv/res_hconf.h b/resolv/res_hconf.h
>>> index b97734d..a3d23f3 100644
>>> --- a/resolv/res_hconf.h
>>> +++ b/resolv/res_hconf.h
>>> @@ -25,6 +25,15 @@
>>>
>>>  struct hconf
>>>  {
>>> +  /* We keep the INITIALIZED member only for backwards compatibility.  New
>>> +     code should just call _res_hconf_init unconditionally.  For this field
>>> +     to be used safely, users must ensure that either (1) a call to
>>> +     _res_hconf_init happens-before any load from INITIALIZED, or (2) an
>>                                ^
>>> +     assignment of zero to INITIALIZED happens-before any load from it, and
>>                                                  ^
>>
>> Those hyphens look wrong.
>
> “happens-before” is third-person singular of the verb
> “happen-before”. It's a technical term widely used while discussing memory
> models.

But writing it without hyphen is standard English and doesn't require
special knowledge.

Andreas.

-- 
Andreas Schwab, SUSE Labs, schwab@suse.de
GPG Key fingerprint = 0196 BAD8 1CE9 1970 F4BE  1748 E4D4 88E3 0EEA B9D7
"And now for something completely different."

Re: [PATCH Fix incorrect double-checked locking related to _res_hconf.initialized. [BZ #20477]

[Florian Weimer]

On 08/18/2016 09:05 AM, Andreas Schwab wrote:
> On Aug 17 2016, Florian Weimer <fweimer@redhat.com> wrote:
>
>> On 08/17/2016 04:51 PM, Andreas Schwab wrote:
>>> On Aug 17 2016, Torvald Riegel <triegel@redhat.com> wrote:
>>>
>>>> diff --git a/resolv/res_hconf.h b/resolv/res_hconf.h
>>>> index b97734d..a3d23f3 100644
>>>> --- a/resolv/res_hconf.h
>>>> +++ b/resolv/res_hconf.h
>>>> @@ -25,6 +25,15 @@
>>>>
>>>>  struct hconf
>>>>  {
>>>> +  /* We keep the INITIALIZED member only for backwards compatibility.  New
>>>> +     code should just call _res_hconf_init unconditionally.  For this field
>>>> +     to be used safely, users must ensure that either (1) a call to
>>>> +     _res_hconf_init happens-before any load from INITIALIZED, or (2) an
>>>                                ^
>>>> +     assignment of zero to INITIALIZED happens-before any load from it, and
>>>                                                  ^
>>>
>>> Those hyphens look wrong.
>>
>> “happens-before” is third-person singular of the verb
>> “happen-before”. It's a technical term widely used while discussing memory
>> models.
>
> But writing it without hyphen is standard English and doesn't require
> special knowledge.

But making sense of this comment requires special knowledge, and writing 
the verb as “happens-before” makes it clear that the memory relation is 
meant, and not some foggy concept of ordering.  It's like 
MAY/SHOULD/RECOMMEND in IETF RFCs in this regard.

Florian

Re: [PATCH Fix incorrect double-checked locking related to _res_hconf.initialized. [BZ #20477]

[Torvald Riegel]

On Thu, 2016-08-18 at 10:05 +0200, Florian Weimer wrote:
> On 08/18/2016 09:05 AM, Andreas Schwab wrote:
> > On Aug 17 2016, Florian Weimer <fweimer@redhat.com> wrote:
> >
> >> On 08/17/2016 04:51 PM, Andreas Schwab wrote:
> >>> On Aug 17 2016, Torvald Riegel <triegel@redhat.com> wrote:
> >>>
> >>>> diff --git a/resolv/res_hconf.h b/resolv/res_hconf.h
> >>>> index b97734d..a3d23f3 100644
> >>>> --- a/resolv/res_hconf.h
> >>>> +++ b/resolv/res_hconf.h
> >>>> @@ -25,6 +25,15 @@
> >>>>
> >>>>  struct hconf
> >>>>  {
> >>>> +  /* We keep the INITIALIZED member only for backwards compatibility.  New
> >>>> +     code should just call _res_hconf_init unconditionally.  For this field
> >>>> +     to be used safely, users must ensure that either (1) a call to
> >>>> +     _res_hconf_init happens-before any load from INITIALIZED, or (2) an
> >>>                                ^
> >>>> +     assignment of zero to INITIALIZED happens-before any load from it, and
> >>>                                                  ^
> >>>
> >>> Those hyphens look wrong.
> >>
> >> “happens-before” is third-person singular of the verb
> >> “happen-before”. It's a technical term widely used while discussing memory
> >> models.
> >
> > But writing it without hyphen is standard English and doesn't require
> > special knowledge.
> 
> But making sense of this comment requires special knowledge, and writing 
> the verb as “happens-before” makes it clear that the memory relation is 
> meant, and not some foggy concept of ordering.  It's like 
> MAY/SHOULD/RECOMMEND in IETF RFCs in this regard.

Yes.  As Florian says, this is meant to refer to the happens-before
relation in the C11 memory model, and not some other notion of ordering.
The formalization of the C11 memory model uses "happens-before", so it
seemed like a good choice.

Nonetheless, I'm open to other ways of referring to it, as long as it's
precise and clearly refers to the memory model; in the end, the glibc
community needs to be happy with how we talk about concurrency.  If
anyone has suggestions, let me know.