Do not fuse locations when normalizing constants for Add and Mul

[jorickert]

Without this PR, the location of the Add/Mul will be incorrectly fused with the location of the constant.

[tungld]

LGTM. Thanks!

[flemairen6]

@jenkins-droid test this please

[AlexandreEichenberger]

LGTM, thanks for adding a lit test, very thorough.

[jorickert]

@AlexandreEichenberger Could you take a look at the CI please? It seems to be stuck.
I think I do not have merge rights either, so someone would need to do this for me

[tungld]

@jorickert The Jenkins machines are under maintenance and they are expected to be back soon this week. We will merge this PR and others as soon as the machines are available.

[AlexandreEichenberger]

@jenkins-droid test this please

[jorickert]

@tungld Are the jenkins runners working again? The status is still Expected — Waiting for status to be reported

[jenkins-droid]

Can one of the admins verify this patch?

[tungld]

@jenkins-droid test this please

[jenkins-droid]

Jenkins Linux amd64 Build #16036 [push] Do not fuse locations wh... started at 20:43

[jenkins-droid]

Jenkins Linux s390x Build #16038 [push] Do not fuse locations wh... started at 21:43

[jenkins-droid]

Jenkins Linux ppc64le Build #15065 [push] Do not fuse locations wh... started at 21:58

[tungld]

@tungld Are the jenkins runners working again? The status is still Expected — Waiting for status to be reported

Merged. Thanks for your contribution @jorickert!

[jenkins-droid]

Jenkins Linux amd64 Build #16036 [push] Do not fuse locations wh... passed after 1 hr 25 min

[jenkins-droid]

Jenkins Linux s390x Build #16038 [push] Do not fuse locations wh... passed after 1 hr 33 min

[jenkins-droid]

Jenkins Linux ppc64le Build #15065 [push] Do not fuse locations wh... aborted after 2 hr 7 min