All of lore.kernel.org
 help / color / mirror / Atom feed
From: joserz@linux.vnet.ibm.com
To: David Gibson <david@gibson.dropbear.id.au>
Cc: Richard Henderson <rth@twiddle.net>,
	qemu-ppc@nongnu.org, qemu-devel@nongnu.org,
	bharata@linux.vnet.ibm.com
Subject: Re: [Qemu-devel] [Qemu-ppc] [PATCH 2/7] target-ppc: Implement unsigned quadword left/right shift and unit tests
Date: Mon, 5 Dec 2016 07:35:39 -0200	[thread overview]
Message-ID: <20161205093539.GA16556@pacoca> (raw)
In-Reply-To: <20161205015639.GB12426@umbus.fritz.box>

On Mon, Dec 05, 2016 at 12:56:39PM +1100, David Gibson wrote:
> On Sat, Dec 03, 2016 at 05:37:27PM -0800, Richard Henderson wrote:
> > On 12/02/2016 09:00 PM, Jose Ricardo Ziviani wrote:
> > > +++ b/include/qemu/host-utils.h
> > > @@ -29,6 +29,33 @@
> > >  #include "qemu/bswap.h"
> > > 
> > >  #ifdef CONFIG_INT128
> > > +static inline void urshift(uint64_t *plow, uint64_t *phigh, uint32_t shift)
> > > +{
> > > +    __uint128_t val = ((__uint128_t)*phigh << 64) | *plow;
> > > +    val >>= (shift & 127);
> > > +    *phigh = val >> 64;
> > > +    *plow = val & 0xffffffffffffffff;
> > > +}
> > > +
> > > +static inline void ulshift(uint64_t *plow, uint64_t *phigh,
> > > +                           uint32_t shift, bool *overflow)
> > > +{
> > > +    __uint128_t val = ((__uint128_t)*phigh << 64) | *plow;
> > > +
> > > +    if (shift == 0) {
> > > +        return;
> > > +    }
> > > +
> > > +    if (shift > 127 || (val >> (128 - (shift & 127))) != 0) {
> > > +        *overflow = true;
> > > +    }
> > > +
> > > +    val <<= (shift & 127);
> > > +
> > > +    *phigh = val >> 64;
> > > +    *plow = val & 0xffffffffffffffff;
> > > +}
> > > +
> > 
> > This belongs in qemu/int128.h, not here.  And certainly not predicated on
> > CONFIG_INT128.
> 
> Is there actually any advantage to the __uint128_t based versions over
> the 64-bit versions?

Nothing special here. It just looks more clear (to me) to shift all 128
bits at once than 2x64. But I agree we won't loose to have just one
function outside CONFIG_INT128.

So, I'll remove these two functions and keep only the other two using
uint64_t types.

Anyway I get a bit confused about int128.h and host-utils.h. I see
functions like divu128 and divs128 that could be in int128.h, since
there is no similar operation in int128.h. Is there any rule about it?

Thank you guys for reviewing it!

> 
> -- 
> David Gibson			| I'll have my music baroque, and my code
> david AT gibson.dropbear.id.au	| minimalist, thank you.  NOT _the_ _other_
> 				| _way_ _around_!
> http://www.ozlabs.org/~dgibson

  reply	other threads:[~2016-12-05  9:35 UTC|newest]

Thread overview: 19+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-12-03  4:59 [Qemu-devel] [PATCH 0/7] POWER9 TCG enablements - BCD functions - final part Jose Ricardo Ziviani
2016-12-03  5:00 ` [Qemu-devel] [PATCH 1/7] target-ppc: Implement bcd_is_valid function Jose Ricardo Ziviani
2016-12-03  5:00 ` [Qemu-devel] [PATCH 2/7] target-ppc: Implement unsigned quadword left/right shift and unit tests Jose Ricardo Ziviani
2016-12-04  1:37   ` Richard Henderson
2016-12-05  1:56     ` David Gibson
2016-12-05  9:35       ` joserz [this message]
2016-12-05 22:59         ` [Qemu-devel] [Qemu-ppc] " David Gibson
2016-12-03  5:00 ` [Qemu-devel] [PATCH 3/7] target-ppc: Implement bcds. instruction Jose Ricardo Ziviani
2016-12-05  3:12   ` David Gibson
2016-12-05  3:15     ` David Gibson
2016-12-03  5:00 ` [Qemu-devel] [PATCH 4/7] target-ppc: Implement bcdus. instruction Jose Ricardo Ziviani
2016-12-05  3:14   ` David Gibson
2016-12-03  5:00 ` [Qemu-devel] [PATCH 5/7] target-ppc: Implement bcdsr. instruction Jose Ricardo Ziviani
2016-12-05  3:19   ` David Gibson
2016-12-05  9:45     ` [Qemu-devel] [Qemu-ppc] " joserz
2016-12-05 18:52     ` joserz
2016-12-05 23:01       ` David Gibson
2016-12-03  5:00 ` [Qemu-devel] [PATCH 6/7] target-ppc: Implement bcdtrunc. instruction Jose Ricardo Ziviani
2016-12-03  5:00 ` [Qemu-devel] [PATCH 7/7] " Jose Ricardo Ziviani

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20161205093539.GA16556@pacoca \
    --to=joserz@linux.vnet.ibm.com \
    --cc=bharata@linux.vnet.ibm.com \
    --cc=david@gibson.dropbear.id.au \
    --cc=qemu-devel@nongnu.org \
    --cc=qemu-ppc@nongnu.org \
    --cc=rth@twiddle.net \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.