Re: sparc64 vs. gcc 4.8

To: Eduardo Horvath <eeh%NetBSD.org@localhost>
Subject: Re: sparc64 vs. gcc 4.8
From: "Erik E. Fair" <fair%netbsd.org@localhost>
Date: Mon, 06 Oct 2014 13:30:41 -0700

Is the divergence in memory ordering semantics and mutex implementation
something which can be remedied within the existing mutex API, or should that
API be changed/adjusted also? Or, put another way, can the different semantics
be virtualized in a reasonable way for MI code, or should we (for best
performance, or just correct operation) make more separately maintained MD
code for sparc?

	Erik <fair%netbsd.org@localhost>

References:
- Re: sparc64 vs. gcc 4.8
  - From: Erik Fair
- Re: sparc64 vs. gcc 4.8
  - From: Eduardo Horvath

Prev by Date: Re: sparc64 vs. gcc 4.8
Next by Date: Re: sparc64 vs. gcc 4.8
Previous by Thread: Re: sparc64 vs. gcc 4.8
Next by Thread: Re: sparc64 vs. gcc 4.8
Indexes:

Home | Main Index | Thread Index | Old Index