From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:53422) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1WFWdA-0002Wu-VY for qemu-devel@nongnu.org; Mon, 17 Feb 2014 17:25:48 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1WFWd0-0006Bd-A6 for qemu-devel@nongnu.org; Mon, 17 Feb 2014 17:25:40 -0500 Received: from cantor2.suse.de ([195.135.220.15]:41713 helo=mx2.suse.de) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1WFWcz-0006Ao-U4 for qemu-devel@nongnu.org; Mon, 17 Feb 2014 17:25:30 -0500 From: =?UTF-8?q?Andreas=20F=C3=A4rber?= Date: Mon, 17 Feb 2014 23:24:58 +0100 Message-Id: <1392675899-21210-49-git-send-email-afaerber@suse.de> In-Reply-To: <1392675899-21210-1-git-send-email-afaerber@suse.de> References: <1392675899-21210-1-git-send-email-afaerber@suse.de> MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable Subject: [Qemu-devel] [PULL v2 48/49] qapi: Refine human printing of sizes List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: qemu-devel@nongnu.org Cc: Michael Roth , Paolo Bonzini , =?UTF-8?q?Andreas=20F=C3=A4rber?= , Luiz Capitulino From: Paolo Bonzini This fixes several bugs or shortcomings of the previous pretty-printer. In particular: * use PRIu64 instead of casting to long long * the exact value is included too * the correct unit of measure (MiB, GiB, etc.) is used. PiB and EiB are added too. * due to an off-by-one error, 512*2^30 was printed as 0.500MiB rather tha= n 512MiB. floor(log2(val)) is equal to 63 - clz(val), while the code used = 64. * The desired specification is %g rather than %f, which always uses three decimals in the current code. However %g would switch to scientific notation when the integer part is >=3D 1000 (e.g. 1000*2^30). To keep th= e code simple, switch to the higher power when the integer part is >=3D 100= 0; overflow is avoided by using frexp instead of clz. Suggested-by: Eric Blake Reviewed-by: Igor Mammedov Reviewed-by: Eric Blake Signed-off-by: Paolo Bonzini Signed-off-by: Andreas F=C3=A4rber --- qapi/string-output-visitor.c | 23 ++++++++++++----------- 1 file changed, 12 insertions(+), 11 deletions(-) diff --git a/qapi/string-output-visitor.c b/qapi/string-output-visitor.c index 67a8798..fb1d2e8 100644 --- a/qapi/string-output-visitor.c +++ b/qapi/string-output-visitor.c @@ -15,6 +15,7 @@ #include "qapi/visitor-impl.h" #include "qapi/qmp/qerror.h" #include "qemu/host-utils.h" +#include =20 struct StringOutputVisitor { @@ -47,30 +48,30 @@ static void print_type_size(Visitor *v, uint64_t *obj= , const char *name, Error **errp) { StringOutputVisitor *sov =3D DO_UPCAST(StringOutputVisitor, visitor,= v); - static const char suffixes[] =3D { 'B', 'K', 'M', 'G', 'T' }; + static const char suffixes[] =3D { 'B', 'K', 'M', 'G', 'T', 'P', 'E'= }; uint64_t div, val; char *out; int i; =20 if (!sov->human) { - out =3D g_strdup_printf("%llu", (long long) *obj); + out =3D g_strdup_printf("%"PRIu64, *obj); string_output_set(sov, out); return; } =20 val =3D *obj; =20 - /* Compute floor(log2(val)). */ - i =3D 64 - clz64(val); - - /* Find the power of 1024 that we'll display as the units. */ - i /=3D 10; - if (i >=3D ARRAY_SIZE(suffixes)) { - i =3D ARRAY_SIZE(suffixes) - 1; - } + /* The exponent (returned in i) minus one gives us + * floor(log2(val * 1024 / 1000). The correction makes us + * switch to the higher power when the integer part is >=3D 1000. + */ + frexp(val / (1000.0 / 1024.0), &i); + i =3D (i - 1) / 10; + assert(i < ARRAY_SIZE(suffixes)); div =3D 1ULL << (i * 10); =20 - out =3D g_strdup_printf("%0.03f%c", (double)val/div, suffixes[i]); + out =3D g_strdup_printf("%"PRIu64" (%0.3g %c%s)", val, + (double)val/div, suffixes[i], i ? "iB" : ""); string_output_set(sov, out); } =20 --=20 1.8.4.5