[m-rev.] for review: set optimization options just once

[Zoltan Somogyi]

2020-10-11 21:06 GMT+11:00 "Julien Fischer" <jfischer at opturion.com>:
> On Sun, 11 Oct 2020, Julien Fischer wrote:
> 
>> On Sun, 11 Oct 2020, Zoltan Somogyi wrote:
>>
>>>  For review by Julien, since it is an attempt to fix a bug he reported,
>>>  Mantis 522. The optimization improvement mentioned in the log message
>>>  has bootchecked, but I haven't tried it out on this code yet.
>>
>> I will try it out after you commit.
> 
> It helps, but it's still not sufficient. 

Yes, I know, and I said so.

> Shifting some of the field
> updates in a separate predicate (as in the diff below) avoids the limit.
> I'll commit that unless there are any objections.

I object. That is a crude workaround, while the attached diff
is a more elegant solution, though it needs to be in the installed
compiler, not just in the workspace. I intend to commit this diff
as soon as a final bootcheck is done.

Zoltan.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Log.common
Type: application/octet-stream
Size: 1984 bytes
Desc: not available
URL: <http://lists.mercurylang.org/archives/reviews/attachments/20201011/1deaa4cf/attachment.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: DIFF.common
Type: application/octet-stream
Size: 16575 bytes
Desc: not available
URL: <http://lists.mercurylang.org/archives/reviews/attachments/20201011/1deaa4cf/attachment-0001.obj>