<feed xmlns='http://www.w3.org/2005/Atom'>
<title>linux-stable.git/arch/mips/lib/multi3.c, branch v4.4.136</title>
<subtitle>Linux kernel stable tree</subtitle>
<link rel='alternate' type='text/html' href='https://git.tavy.me/linux-stable.git/'/>
<entry>
<title>MIPS: Implement __multi3 for GCC7 MIPS64r6 builds</title>
<updated>2018-03-03T09:19:46+00:00</updated>
<author>
<name>James Hogan</name>
<email>jhogan@kernel.org</email>
</author>
<published>2017-12-07T07:20:46+00:00</published>
<link rel='alternate' type='text/html' href='https://git.tavy.me/linux-stable.git/commit/?id=cacdb0162489b42f6b46e132b4d08860aa9a07c9'/>
<id>cacdb0162489b42f6b46e132b4d08860aa9a07c9</id>
<content type='text'>
commit ebabcf17bcd7ce968b1631ebe08236275698f39b upstream.

GCC7 is a bit too eager to generate suboptimal __multi3 calls (128bit
multiply with 128bit result) for MIPS64r6 builds, even in code which
doesn't explicitly use 128bit types, such as the following:

unsigned long func(unsigned long a, unsigned long b)
{
	return a &gt; (~0UL) / b;
}

Which GCC rearanges to:

return (unsigned __int128)a * (unsigned __int128)b &gt; 0xffffffffffffffff;

Therefore implement __multi3, but only for MIPS64r6 with GCC7 as under
normal circumstances we wouldn't expect any calls to __multi3 to be
generated from kernel code.

Reported-by: Thomas Petazzoni &lt;thomas.petazzoni@free-electrons.com&gt;
Signed-off-by: James Hogan &lt;jhogan@kernel.org&gt;
Tested-by: Waldemar Brodkorb &lt;wbx@openadk.org&gt;
Cc: Ralf Baechle &lt;ralf@linux-mips.org&gt;
Cc: Maciej W. Rozycki &lt;macro@mips.com&gt;
Cc: Matthew Fortune &lt;matthew.fortune@mips.com&gt;
Cc: Florian Fainelli &lt;florian@openwrt.org&gt;
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/17890/
Cc: Guenter Roeck &lt;linux@roeck-us.net&gt;
Signed-off-by: Greg Kroah-Hartman &lt;gregkh@linuxfoundation.org&gt;

</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
commit ebabcf17bcd7ce968b1631ebe08236275698f39b upstream.

GCC7 is a bit too eager to generate suboptimal __multi3 calls (128bit
multiply with 128bit result) for MIPS64r6 builds, even in code which
doesn't explicitly use 128bit types, such as the following:

unsigned long func(unsigned long a, unsigned long b)
{
	return a &gt; (~0UL) / b;
}

Which GCC rearanges to:

return (unsigned __int128)a * (unsigned __int128)b &gt; 0xffffffffffffffff;

Therefore implement __multi3, but only for MIPS64r6 with GCC7 as under
normal circumstances we wouldn't expect any calls to __multi3 to be
generated from kernel code.

Reported-by: Thomas Petazzoni &lt;thomas.petazzoni@free-electrons.com&gt;
Signed-off-by: James Hogan &lt;jhogan@kernel.org&gt;
Tested-by: Waldemar Brodkorb &lt;wbx@openadk.org&gt;
Cc: Ralf Baechle &lt;ralf@linux-mips.org&gt;
Cc: Maciej W. Rozycki &lt;macro@mips.com&gt;
Cc: Matthew Fortune &lt;matthew.fortune@mips.com&gt;
Cc: Florian Fainelli &lt;florian@openwrt.org&gt;
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/17890/
Cc: Guenter Roeck &lt;linux@roeck-us.net&gt;
Signed-off-by: Greg Kroah-Hartman &lt;gregkh@linuxfoundation.org&gt;

</pre>
</div>
</content>
</entry>
</feed>
