diff mbox series

[wwwdocs,v2] gcc-15/changes: Mention the new -mveclibabi=aocl option in the IA-32/x86-64 section

Message ID Z8rkQeh2-FC3W5vO@localhost.localdomain
State New
Headers
Series [wwwdocs,v2] gcc-15/changes: Mention the new -mveclibabi=aocl option in the IA-32/x86-64 section |

Checks

Context Check Description
linaro-tcwg-bot/tcwg_simplebootstrap_build--master-arm-bootstrap fail Patch failed to apply
linaro-tcwg-bot/tcwg_gcc_build--master-arm fail Patch failed to apply
linaro-tcwg-bot/tcwg_gcc_build--master-aarch64 fail Patch failed to apply
linaro-tcwg-bot/tcwg_simplebootstrap_build--master-aarch64-bootstrap fail Patch failed to apply

Commit Message

Filip Kastl March 7, 2025, 12:19 p.m. UTC 
  Hi Gerald,

On Thu 2025-03-06 18:38:12, Gerald Pfeifer wrote:
> On Thu, 6 Mar 2025, Filip Kastl wrote:
> > I would like to gently ping this.
> 
> Ooops, sorry.
> 
> > > +  <li>GCC now supports generating vectorized math calls to the math library
> > > +    from AMD Optimizing CPU Libraries (AOCL LibM). This option is available
> > > +    through the <code>-mveclibabi=aocl</code> compiler switch. GCC continues to
> > > +    support generating calls to AMD Core Math Library (ACML). However, that
> > > +    library is end-of-life and AOCL offers many more vectorized functions.
> > > +  </li>
> 
> This is fine, though I suggest to work out a bit more clearly what the 
> default is. I guess it's neither of these two, unless... ?
> 
> And one specific suggestion (which you may opt to implement or not): How 
> about "GCC still supports generating" to avoid having three verb forms in
> sequence (which made my head hurt a bit ;-)?
> 
> Gerald

Here is a second version of the patch.  Thanks for the suggestions.  I
addressed them both.  Do you think this is better?

Thanks,
Filip Kastl


-- 8< --


---
 htdocs/gcc-15/changes.html | 8 ++++++++
 1 file changed, 8 insertions(+)

Comments

Gerald Pfeifer March 7, 2025, 3:36 p.m. UTC | #1 
On Fri, 7 Mar 2025, Filip Kastl wrote:
> Here is a second version of the patch.  Thanks for the suggestions.  I
> addressed them both.  Do you think this is better?

Loving it. :)

Thank you,
Gerald
Filip Kastl March 8, 2025, 8:17 p.m. UTC | #2 
On Fri 2025-03-07 16:36:17, Gerald Pfeifer wrote:
> On Fri, 7 Mar 2025, Filip Kastl wrote:
> > Here is a second version of the patch.  Thanks for the suggestions.  I
> > addressed them both.  Do you think this is better?
> 
> Loving it. :)
> 
> Thank you,
> Gerald

Ok, pushed.  Thanks!

Filip
diff mbox series

Patch

diff --git a/htdocs/gcc-15/changes.html b/htdocs/gcc-15/changes.html
index c1116899..bc071501 100644
--- a/htdocs/gcc-15/changes.html
+++ b/htdocs/gcc-15/changes.html
@@ -592,6 +592,14 @@  asm (".text; %cc0: mov %cc2, %%r0; .previous;"
       <code>-mavx512pf</code>, <code>-mprefetchwt1</code>,
       <code>-mtune=knl</code>, and <code>-mtune=knm</code> compiler switches.
   </li>
+  <li>With the <code>-mveclibabi</code> compiler switch GCC is able to generate
+    vectorized calls to external libraries. GCC 15 newly supports generating
+    vectorized math calls to the math library from AMD Optimizing CPU Libraries
+    (AOCL LibM). This option is available through
+    <code>-mveclibabi=aocl</code>. GCC still supports generating calls to AMD
+    Core Math Library (ACML). However, that library is end-of-life and AOCL
+    offers many more vectorized functions.
+  </li>
 </ul>
 
 <!-- <h3 id="mips">MIPS</h3> -->