[PATCH] Document experimental status of odr

[PATCH] Document experimental status of odr

[Tom de Vries]

Hi,

Add documentation of the experimental status of the odr optimization in dwz.1.

Any comments?

Thanks,
- Tom

Document experimental status of odr

2021-02-25  Tom de Vries  <tdevries@suse.de>

	PR dwz/27401
	* dwz.1: Document experimental status of odr.

---
 dwz.1 | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/dwz.1 b/dwz.1
index e9117b2..19dc814 100644
--- a/dwz.1
+++ b/dwz.1
@@ -112,11 +112,13 @@ corresponding forms, instead of the GNU extension \fI.gnu_debugaltlink\fR
 and corresponding forms.
 .TP
 .B \-\-odr / \-\-no-odr
+.B Experimental.
 Enable/disable One-Definition-Rule optimization for C++ compilation units.
 This optimization causes struct/union/class DIEs with the same name to be
 considered equal.  This has the effect that DIEs referring to distinct DIEs
 representing the same type (like f.i. pointer type DIEs) are considered equal,
-and may be deduplicated.
+and may be deduplicated.  The status of this optimization is experimental.
+It's disabled in low-mem mode.
 Disabled by default.
 .TP
 .B \-\-odr-mode=<basic|link>

Re: [PATCH] Document experimental status of odr

[Mark Wielaard]

Hi Tom,

On Thu, Feb 25, 2021 at 05:17:31PM +0100, Tom de Vries wrote:
> Add documentation of the experimental status of the odr optimization in dwz.1.
> 
> Any comments?
>
> 
> 	PR dwz/27401
> 	* dwz.1: Document experimental status of odr.

Looks good to me.

BTW. Do you already know when would you consider it no longer
experimental? Are there still fundemental bugs/design issues or is it
now just having people experiment with it and finding bugs?

Cheers,

Mark

Re: [PATCH] Document experimental status of odr

[Tom de Vries]

On 2/26/21 2:04 PM, Mark Wielaard wrote:
> Hi Tom,
> 
> On Thu, Feb 25, 2021 at 05:17:31PM +0100, Tom de Vries wrote:
>> Add documentation of the experimental status of the odr optimization in dwz.1.
>>
>> Any comments?
>>
>>
>> 	PR dwz/27401
>> 	* dwz.1: Document experimental status of odr.
> 
> Looks good to me.
> 

Thanks for the review.

> BTW. Do you already know when would you consider it no longer
> experimental? Are there still fundemental bugs/design issues

With the current state of trunk + recently submitted odr fixes there are
no longer any known bugs in odr.

The only design issue that I can think of is that it's disabled for
low-mem (listed in set_die_odr_state as todo).

> or is it
> now just having people experiment with it and finding bugs?

That as well.

[ FWIW, I've unofficially labeled a recent commit as 0.14-rc1 and am in
the process of bringing that + submitted odr fixes to tumbleweed, with
odr enabled by default.  So, I expect some feedback there ... ;) ]

To come back to your question, the label experimental is to warn the
user that it can be used, but expect problems.  So I'd say once we fix
the initially reported problems and figure out the low-mem thingy, we
can drop the experimental label.

Thanks,
- Tom