From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <SRS0=WP31=5I=gmail.com=goldstein.w.n@sourceware.org>
Received: from mail-ed1-x52b.google.com (mail-ed1-x52b.google.com [IPv6:2a00:1450:4864:20::52b])
	by sourceware.org (Postfix) with ESMTPS id A3EFA3858C83
	for <libc-alpha@sourceware.org>; Wed, 11 Jan 2023 19:56:39 +0000 (GMT)
DMARC-Filter: OpenDMARC Filter v1.4.2 sourceware.org A3EFA3858C83
Authentication-Results: sourceware.org; dmarc=pass (p=none dis=none) header.from=gmail.com
Authentication-Results: sourceware.org; spf=pass smtp.mailfrom=gmail.com
Received: by mail-ed1-x52b.google.com with SMTP id x10so21044704edd.10
        for <libc-alpha@sourceware.org>; Wed, 11 Jan 2023 11:56:39 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed;
        d=gmail.com; s=20210112;
        h=cc:to:subject:message-id:date:from:in-reply-to:references
         :mime-version:from:to:cc:subject:date:message-id:reply-to;
        bh=zsNt6wWd0BYxihaQ88r0uHwV06rt2yS/oTZsKW/wk+U=;
        b=GN0/qPLoPa79poeqb2CAXDWPi9/dRDC8TDsCaAP4v7n8kim6T2/tqUP2fVW5lmgN/L
         MEzWN2pR0AKIFXklElUmRDFsK8XlImgiVW5+3NsVtGVpHm9euhEo/swl8twO8kYgJW2s
         l9mB9DjRKVz2Et3tYtoheYHu6vPM8W8IY7UlmXHqWR5iOPollifgBqEg1TYJApriy31X
         Iazhp8ABb/CceUtfVOksZ+K8KGv2PoTD8bjo6XVhX5JqBDMpllBwIVr5saATVsWLhi5r
         jKX6aXalMqnjwUHceAbssMnfw6NmiiBt2LAoxAAUiqaeLL+xrOhBJ5x5ytBR29ySPrUK
         VZSQ==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed;
        d=1e100.net; s=20210112;
        h=cc:to:subject:message-id:date:from:in-reply-to:references
         :mime-version:x-gm-message-state:from:to:cc:subject:date:message-id
         :reply-to;
        bh=zsNt6wWd0BYxihaQ88r0uHwV06rt2yS/oTZsKW/wk+U=;
        b=CGuTa1GZnn/eLtaB5Qm+W21miiw1YkUZisL1zrDijm75DWvSQGrHkr9Jki5OfEeLda
         N4EpgbWrqXMSgQ8BChwS4v8kJ92TiVDy78d8LxHR9+2qnOze19jMqVomWbYnOul6vdqX
         ja1OJDm9l+NxbbRSnt5llUd2j3Iu85fzM/YW4Zny6rmMJmTJr6Y29wXT6F9Xcq43OJuO
         +X6Fo4IQwRSIaUUWCsOrBUzZ1xSuVc3xcuUFvgt3VhbeeGzJ9YBcO9zNKXYlyWJxVGtl
         1zU3lDsKJfl5JFOlCf1iz7Kg0pdACkqWf5rrNtSUp+Oz1mnnH2fipgQgyYx5KPXKo1cM
         J6ww==
X-Gm-Message-State: AFqh2kpFO4utHJvbKCXK5treAGqJjR4uJBwOSNVUW1M6aaQc8nAN02tL
	y9mWo1EvUze9wuepi1MRKA7iGybzp3Z9et/oMGkRs3v2
X-Google-Smtp-Source: AMrXdXtLBQbPjNaB9HrrF4AnrkMnD9DQIbNIa4SMkJOyiMRPM0hDOQuErl+OpEKzw9LZuZAnTR0sylYYA+sSDa267Mg=
X-Received: by 2002:a05:6402:2408:b0:47b:2514:f2e7 with SMTP id
 t8-20020a056402240800b0047b2514f2e7mr9116460eda.142.1673466998353; Wed, 11
 Jan 2023 11:56:38 -0800 (PST)
MIME-Version: 1.0
References: <20230110210106.1457686-1-adhemerval.zanella@linaro.org>
 <20230110210106.1457686-5-adhemerval.zanella@linaro.org> <f60482cf-e782-bff9-0793-a0e1674b2fdd@linaro.org>
 <7b60bc6b-9e33-4bb4-0a64-3b28ca1302fe@linaro.org>
In-Reply-To: <7b60bc6b-9e33-4bb4-0a64-3b28ca1302fe@linaro.org>
From: Noah Goldstein <goldstein.w.n@gmail.com>
Date: Wed, 11 Jan 2023 11:56:26 -0800
Message-ID: <CAFUsyf+=6h=vrK1k4r-HDWMxYM=gwUkDPz5N7hXt_nFQ6Ep4Mw@mail.gmail.com>
Subject: Re: [PATCH v6 04/17] Add string vectorized find and detection functions
To: Adhemerval Zanella Netto <adhemerval.zanella@linaro.org>
Cc: Richard Henderson <richard.henderson@linaro.org>, libc-alpha@sourceware.org
Content-Type: text/plain; charset="UTF-8"
X-Spam-Status: No, score=-3.6 required=5.0 tests=BAYES_00,DKIM_SIGNED,DKIM_VALID,DKIM_VALID_AU,DKIM_VALID_EF,FREEMAIL_FROM,RCVD_IN_DNSWL_NONE,SPF_HELO_NONE,SPF_PASS,TXREP autolearn=ham autolearn_force=no version=3.4.6
X-Spam-Checker-Version: SpamAssassin 3.4.6 (2021-04-09) on server2.sourceware.org
List-Id: <libc-alpha.sourceware.org>

On Wed, Jan 11, 2023 at 5:42 AM Adhemerval Zanella Netto
<adhemerval.zanella@linaro.org> wrote:
>
>
>
> On 10/01/23 23:25, Richard Henderson wrote:
> > On 1/10/23 13:00, Adhemerval Zanella wrote:
> >> --- a/sysdeps/generic/string-maskoff.h
> >> +++ b/sysdeps/generic/string-maskoff.h
> >> @@ -23,7 +23,6 @@
> >>   #include <limits.h>
> >>   #include <stdint.h>
> >>   #include <string-optype.h>
> >> -#include <string-fza.h>
> >>     /* Provide a mask based on the pointer alignment that sets up non-zero
> >>      bytes before the beginning of the word.  It is used to mask off
> >> @@ -44,12 +43,12 @@ create_mask (uintptr_t i)
> >>   /* Return the mask WORD shifted based on S_INT address value, to ignore
> >>      values not presented in the aligned word read.  */
> >>   static __always_inline op_t
> >> -check_mask (op_t word, uintptr_t s_int)
> >> +check_mask (op_t word, uintptr_t s)
> >>   {
> >>     if (__BYTE_ORDER == __LITTLE_ENDIAN)
> >> -    return word >> (CHAR_BIT * (s_int % sizeof (s_int)));
> >> +    return word >> (CHAR_BIT * (s % sizeof (op_t)));
> >>     else
> >> -    return word << (CHAR_BIT * (s_int % sizeof (s_int)));
> >> +    return word << (CHAR_BIT * (s % sizeof (op_t)));
> >>   }
> >
> > Fold this into patch 3, if you want it.
>
> Ack.
>
> >
> >> @@ -79,7 +78,7 @@ highbit_mask (op_t m)
> >>   static __always_inline op_t *
> >>   word_containing (char const *p)
> >>   {
> >> -  return (op_t *) ((uintptr_t) p & -sizeof(p));
> >> +  return (op_t *) ((op_t) p & -sizeof(p));
> >
> > Casting p to op_t is wrong.
>
> Indeed, I will revert to uintptr_t.

If sizeof (op_t) > sizeof(uintptr_t) won't this be able to
leave us with a misaligned read + potential fault?