From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932472AbcARWVO (ORCPT ); Mon, 18 Jan 2016 17:21:14 -0500 Received: from mail-wm0-f46.google.com ([74.125.82.46]:35826 "EHLO mail-wm0-f46.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S932372AbcARWVK convert rfc822-to-8bit (ORCPT ); Mon, 18 Jan 2016 17:21:10 -0500 From: Rasmus Villemoes To: Andy Shevchenko Cc: Robert Elliott , Matt Fleming , Andrew Morton , Thomas Gleixner , Ingo Molnar , "H. Peter Anvin" , linux-kernel@vger.kernel.org, Brian Norris , Hariprasad S Subject: Re: [PATCH v2 09/11] lib/vsprintf: use precision field with %pl[From[To]] Organization: D03 References: <1452810221-116505-1-git-send-email-andriy.shevchenko@linux.intel.com> <1452810221-116505-10-git-send-email-andriy.shevchenko@linux.intel.com> X-Hashcash: 1:20:160118:elliott@hpe.com::ADozalECYwJq2gWN:000ZFN X-Hashcash: 1:20:160118:andriy.shevchenko@linux.intel.com::oyy+mX2Fh0e3nRei:00000000000000000000000000000QQn X-Hashcash: 1:20:160118:matt@codeblueprint.co.uk::npA5xpscAIfJJZWA:00000000000000000000000000000000000000zWU X-Hashcash: 1:20:160118:hpa@zytor.com::Aa/QhJJd1MT+gvGf:00001Mll X-Hashcash: 1:20:160118:mingo@redhat.com::xHoEe1yWI8gh05vu:01sdW X-Hashcash: 1:20:160118:akpm@linux-foundation.org::db4RWxt2DDyxdYpT:0000000000000000000000000000000000003Pf9 X-Hashcash: 1:20:160118:tglx@linutronix.de::AXbRUr1dX8wK2i5c:00000000000000000000000000000000000000000003qlD X-Hashcash: 1:20:160118:linux-kernel@vger.kernel.org::P13phladJRSs/JZ3:0000000000000000000000000000000004V38 X-Hashcash: 1:20:160118:hariprasad@chelsio.com::L7aKjwS/WV1VJ10r:00000000000000000000000000000000000000050b/ X-Hashcash: 1:20:160118:computersforpeace@gmail.com::6nuXEi+fAEPcFkU7:00000000000000000000000000000000005rSf Date: Mon, 18 Jan 2016 23:21:07 +0100 In-Reply-To: <1452810221-116505-10-git-send-email-andriy.shevchenko@linux.intel.com> (Andy Shevchenko's message of "Fri, 15 Jan 2016 00:23:39 +0200") Message-ID: <87egdeedgs.fsf@rasmusvillemoes.dk> User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/24.3 (gnu/linux) MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 8BIT Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Thu, Jan 14 2016, Andy Shevchenko wrote: > The precision field defines how many digits after period ('.') will be > printed. Yuck again. Didn't you notice gcc barking at you? warning: precision used with ‘%p’ gnu_printf format [-Wformat=] So I don't think this will ever be used. > Signed-off-by: Andy Shevchenko > --- > Documentation/printk-formats.txt | 7 +++++++ > lib/test_printf.c | 29 +++++++++++++++++++++++++++++ > lib/vsprintf.c | 20 +++++++++++++++++++- > 3 files changed, 55 insertions(+), 1 deletion(-) > > diff --git a/Documentation/printk-formats.txt b/Documentation/printk-formats.txt > index 3b41899..534407c 100644 > --- a/Documentation/printk-formats.txt > +++ b/Documentation/printk-formats.txt > @@ -295,11 +295,18 @@ unsigned long long value in human-readable form (with IEC binary prefix): > %pl 1048575 B (1048575) > %plKM 1023 KiB (1048575) > %pl 1 MiB (1048576) > + %.1plKM 1046527.9 MiB (1097364144125) > %plG 2 TiB (2199023255552) > > For printing given unsigned long long value in human-readable form with > automatically choosen IEC binary prefix. > > + Precision, if passed, defines how many digits will be printed after > + period ('.'). In this case the biggest possible unit is used in the > + given range [From[To]]. The printed value isn't rounded anyhow. > + > + For other use cases consider to call string_get_size(). > + > Passed by reference. > > bitmap and its derivatives such as cpumask and nodemask: > diff --git a/lib/test_printf.c b/lib/test_printf.c > index ae35885..4a2e30b 100644 > --- a/lib/test_printf.c > +++ b/lib/test_printf.c > @@ -435,6 +435,30 @@ human_readable(void) > { .v = 1048575ULL, .r = "1023 KiB" }, > { .v = 1047552ULL, .r = "1023 KiB" }, > }; > + struct hr_test_data htdpl0KM[] = { > + { .v = 1097364144128ULL, .r = "1046528 MiB" }, > + { .v = 1097364144125ULL, .r = "1046527 MiB" }, > + { .v = 1048578ULL, .r = "1 MiB" }, > + { .v = 1048576ULL, .r = "1 MiB" }, > + { .v = 1048575ULL, .r = "1023 KiB" }, > + { .v = 1047552ULL, .r = "1023 KiB" }, > + }; > + struct hr_test_data htdpl1KM[] = { > + { .v = 1097364144128ULL, .r = "1046528.0 MiB" }, > + { .v = 1097364144125ULL, .r = "1046527.9 MiB" }, > + { .v = 1048578ULL, .r = "1.0 MiB" }, > + { .v = 1048576ULL, .r = "1.0 MiB" }, > + { .v = 1048575ULL, .r = "1023.9 KiB" }, > + { .v = 1047552ULL, .r = "1023.0 KiB" }, > + }; > + struct hr_test_data htdpl3KM[] = { > + { .v = 1097364144128ULL, .r = "1046528.000 MiB" }, > + { .v = 1097364144125ULL, .r = "1046527.999 MiB" }, > + { .v = 1048578ULL, .r = "1.000 MiB" }, > + { .v = 1048576ULL, .r = "1.000 MiB" }, > + { .v = 1048575ULL, .r = "1023.999 KiB" }, > + { .v = 1047552ULL, .r = "1023.000 KiB" }, > + }; > struct hr_test_array { > const char *fmt; > struct hr_test_data *data; > @@ -446,6 +470,11 @@ human_readable(void) > { "%plKM", htdplKM, ARRAY_SIZE(htdplKM) }, > /* No reversed %pl[To[From]] order */ > { "%plMK", htdplM, ARRAY_SIZE(htdplM) }, > + { "%.0plKM", htdpl0KM, ARRAY_SIZE(htdpl0KM) }, > + { "%.1plKM", htdpl1KM, ARRAY_SIZE(htdpl1KM) }, > + { "%.3plKM", htdpl3KM, ARRAY_SIZE(htdpl3KM) }, > + /* For now precision is in range [0..3] */ > + { "%.4plKM", htdplKM, ARRAY_SIZE(htdplKM) }, > }; > unsigned int i, j; > > diff --git a/lib/vsprintf.c b/lib/vsprintf.c > index fbcb221..2dd86c9 100644 > --- a/lib/vsprintf.c > +++ b/lib/vsprintf.c > @@ -1327,7 +1327,9 @@ char *human_readable(char *buf, char *end, const void *addr, > to = i; > } > > - if (value) > + if (spec.precision >= 0 && spec.precision <= 3 && value) > + index = (fls64(value) - 1) / 10; > + else if (value) > index = __ffs64(value) / 10; > else > index = 0; > @@ -1336,6 +1338,22 @@ char *human_readable(char *buf, char *end, const void *addr, > /* Print numeric part */ > buf = number(buf, end, value >> (index * 10), default_dec_spec); > > + /* Print precision part if asked */ > + if (spec.precision > 0 && spec.precision <= 3 && index) { > + unsigned int remainder; > + > + remainder = (value >> (index - 1) * 10) % 1024; > + for (i = 0; i < spec.precision; i++) > + remainder *= 10; > + remainder >>= 10; > + > + /* Period... */ > + buf = put_one_char(buf, end, '.'); > + > + /* ...and remainder part */ > + buf = number(buf, end, remainder, spec); > + } > + > /* Space between numeric part and units */ > buf = put_one_char(buf, end, ' ');