From patchwork Tue Sep 5 18:21:15 2017 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Ulrich Weigand X-Patchwork-Id: 22639 Received: (qmail 56710 invoked by alias); 5 Sep 2017 18:21:30 -0000 Mailing-List: contact gdb-patches-help@sourceware.org; run by ezmlm Precedence: bulk List-Id: List-Unsubscribe: List-Subscribe: List-Archive: List-Post: List-Help: , Sender: gdb-patches-owner@sourceware.org Delivered-To: mailing list gdb-patches@sourceware.org Received: (qmail 56669 invoked by uid 89); 5 Sep 2017 18:21:30 -0000 Authentication-Results: sourceware.org; auth=none X-Virus-Found: No X-Spam-SWARE-Status: No, score=-11.8 required=5.0 tests=AWL, BAYES_00, GIT_PATCH_2, GIT_PATCH_3, KAM_ASCII_DIVIDERS, RCVD_IN_DNSWL_LOW, SPF_PASS autolearn=ham version=3.3.2 spammy= X-HELO: mx0a-001b2d01.pphosted.com Received: from mx0a-001b2d01.pphosted.com (HELO mx0a-001b2d01.pphosted.com) (148.163.156.1) by sourceware.org (qpsmtpd/0.93/v0.84-503-g423c35a) with ESMTP; Tue, 05 Sep 2017 18:21:23 +0000 Received: from pps.filterd (m0098394.ppops.net [127.0.0.1]) by mx0a-001b2d01.pphosted.com (8.16.0.21/8.16.0.21) with SMTP id v85IKKRd085119 for ; Tue, 5 Sep 2017 14:21:22 -0400 Received: from e06smtp12.uk.ibm.com (e06smtp12.uk.ibm.com [195.75.94.108]) by mx0a-001b2d01.pphosted.com with ESMTP id 2csvh0f0uy-1 (version=TLSv1.2 cipher=AES256-SHA bits=256 verify=NOT) for ; Tue, 05 Sep 2017 14:21:21 -0400 Received: from localhost by e06smtp12.uk.ibm.com with IBM ESMTP SMTP Gateway: Authorized Use Only! Violators will be prosecuted for from ; Tue, 5 Sep 2017 19:21:19 +0100 Received: from b06cxnps4075.portsmouth.uk.ibm.com (9.149.109.197) by e06smtp12.uk.ibm.com (192.168.101.142) with IBM ESMTP SMTP Gateway: Authorized Use Only! Violators will be prosecuted; Tue, 5 Sep 2017 19:21:16 +0100 Received: from d06av24.portsmouth.uk.ibm.com (mk.ibm.com [9.149.105.60]) by b06cxnps4075.portsmouth.uk.ibm.com (8.14.9/8.14.9/NCO v10.0) with ESMTP id v85ILFTO11665578 for ; Tue, 5 Sep 2017 18:21:15 GMT Received: from d06av24.portsmouth.uk.ibm.com (unknown [127.0.0.1]) by IMSVA (Postfix) with ESMTP id 0EAF24203F for ; Tue, 5 Sep 2017 19:17:47 +0100 (BST) Received: from d06av24.portsmouth.uk.ibm.com (unknown [127.0.0.1]) by IMSVA (Postfix) with ESMTP id E471C42041 for ; Tue, 5 Sep 2017 19:17:46 +0100 (BST) Received: from oc3748833570.ibm.com (unknown [9.164.150.234]) by d06av24.portsmouth.uk.ibm.com (Postfix) with ESMTP for ; Tue, 5 Sep 2017 19:17:46 +0100 (BST) Received: by oc3748833570.ibm.com (Postfix, from userid 1000) id 4163DD8086F; Tue, 5 Sep 2017 20:21:15 +0200 (CEST) Subject: [RFC][09/19] Target FP: Introduce target-float.{c,h} To: gdb-patches@sourceware.org Date: Tue, 5 Sep 2017 20:21:15 +0200 (CEST) From: "Ulrich Weigand" MIME-Version: 1.0 X-TM-AS-GCONF: 00 x-cbid: 17090518-0008-0000-0000-00000492CDCD X-IBM-AV-DETECTION: SAVI=unused REMOTE=unused XFE=unused x-cbparentid: 17090518-0009-0000-0000-00001E23342D Message-Id: <20170905182115.4163DD8086F@oc3748833570.ibm.com> X-Proofpoint-Virus-Version: vendor=fsecure engine=2.50.10432:, , definitions=2017-09-05_07:, , signatures=0 X-Proofpoint-Spam-Details: rule=outbound_notspam policy=outbound score=0 spamscore=0 suspectscore=43 malwarescore=0 phishscore=0 adultscore=0 bulkscore=0 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.0.1-1707230000 definitions=main-1709050267 [RFC][09/19] Target FP: Introduce target-float.{c,h} This patch introduces the new set of target floating-point handling routines in target-float.{c,h}. In the end, the intention is that this file will contain support for all operations in target FP format, fully replacing both the current doublest.{c,h} and dfp.{c,h}. To begin with, this patch only adds a target_float_is_zero routine, which handles the equivalent of decimal_is_zero for both binary and decimal FP. For the binary case, to avoid conversion to DOUBLEST, this is implemented using the floatformat_classify routine. However, it turns out that floatformat_classify actually has a bug (it was not used to check for zero before), so this is fixed as well. The new routine is used in both value_logical_not and valpy_nonzero. There is one extra twist: the code previously used value_as_double to convert to DOUBLEST and then compare against zero. That routine performs an extra task: it detects invalid floating-point values and raises an error. In any place where value_as_double is removed in favor of some target-float.c routine, we need to replace that check. To keep this check centralized in one place, I've added a new routine is_floating_value, which returns a boolean determining whether a value's type is floating point (binary or decimal), and if so, also performs the validity check. Since we need to check whether a value is FP before calling any of the target-float routines anyway, this seems a good place to add the check without much code size overhead. In some places where we only want to check for floating-point types and not perform a validity check (e.g. for the *output* of an operation), we can use the new is_floating_type routine (in gdbarch) instead. The validity check itself is done by a new target_float_is_valid routine in target-float, encapsulating floatformat_is_valid. Bye, Ulrich ChangeLog: * Makefile.c (SFILES): Add target-float.c. (HFILES_NO_SRCDIR): Add target-float.h. (COMMON_OBS): Add target-float.o. * target-float.h: New file. * target-float.c: New file. * doublest.c (floatformat_classify): Fix detection of float_zero. * gdbtypes.c (is_floating_type): New function. * gdbtypes.h (is_floating_type): Add prototype. * value.c: Do not include "floatformat.h". (unpack_double): Use target_float_is_valid. (is_floating_value): New function. * value.h (is_floating_value): Add prototype- * valarith.c: Include "target-float.h". (value_logical_not): Use target_float_is_zero. * python/py-value.c: Include "target-float.h". (valpy_nonzero): Use target_float_is_zero. Index: binutils-gdb/gdb/Makefile.in =================================================================== --- binutils-gdb.orig/gdb/Makefile.in +++ binutils-gdb/gdb/Makefile.in @@ -1214,6 +1214,7 @@ SFILES = \ tracepoint.c \ trad-frame.c \ tramp-frame.c \ + target-float.c \ typeprint.c \ ui-file.c \ ui-file.h \ @@ -1482,6 +1483,7 @@ HFILES_NO_SRCDIR = \ tracefile.h \ tracepoint.h \ trad-frame.h \ + target-float.h \ tramp-frame.h \ typeprint.h \ ui-file.h \ @@ -1832,6 +1834,7 @@ COMMON_OBS = $(DEPFILES) $(CONFIG_OBS) $ top.o \ trad-frame.o \ tramp-frame.o \ + target-float.o \ typeprint.o \ ui-file.o \ ui-out.o \ Index: binutils-gdb/gdb/target-float.c =================================================================== --- /dev/null +++ binutils-gdb/gdb/target-float.c @@ -0,0 +1,62 @@ +/* Floating point routines for GDB, the GNU debugger. + + Copyright (C) 2017 Free Software Foundation, Inc. + + This file is part of GDB. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 3 of the License, or + (at your option) any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program. If not, see . */ + +#include "defs.h" +#include "dfp.h" +#include "doublest.h" +#include "gdbtypes.h" +#include "floatformat.h" +#include "target-float.h" + + +/* Typed floating-point routines. These routines operate on floating-point + values in target format, represented by a byte buffer interpreted as a + "struct type", which may be either a binary or decimal floating-point + type (TYPE_CODE_FLT or TYPE_CODE_DECFLOAT). */ + +/* Return whether the byte-stream ADDR holds a valid value of + floating-point type TYPE. */ +bool +target_float_is_valid (const gdb_byte *addr, const struct type *type) +{ + if (TYPE_CODE (type) == TYPE_CODE_FLT) + return floatformat_is_valid (floatformat_from_type (type), addr); + + if (TYPE_CODE (type) == TYPE_CODE_DECFLOAT) + return true; + + gdb_assert_not_reached ("unexpected type code"); +} + +/* Return whether the byte-stream ADDR, interpreted as floating-point + type TYPE, is numerically equal to zero (of either sign). */ +bool +target_float_is_zero (const gdb_byte *addr, const struct type *type) +{ + if (TYPE_CODE (type) == TYPE_CODE_FLT) + return (floatformat_classify (floatformat_from_type (type), addr) + == float_zero); + + if (TYPE_CODE (type) == TYPE_CODE_DECFLOAT) + return decimal_is_zero (addr, TYPE_LENGTH (type), + gdbarch_byte_order (get_type_arch (type))); + + gdb_assert_not_reached ("unexpected type code"); +} + Index: binutils-gdb/gdb/target-float.h =================================================================== --- /dev/null +++ binutils-gdb/gdb/target-float.h @@ -0,0 +1,28 @@ +/* Floating point definitions for GDB. + + Copyright (C) 1986-2017 Free Software Foundation, Inc. + + This file is part of GDB. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 3 of the License, or + (at your option) any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program. If not, see . */ + +#ifndef TYPED_FLOAT_H +#define TYPED_FLOAT_H + +extern bool target_float_is_valid (const gdb_byte *addr, + const struct type *type); +extern bool target_float_is_zero (const gdb_byte *addr, + const struct type *type); + +#endif Index: binutils-gdb/gdb/doublest.c =================================================================== --- binutils-gdb.orig/gdb/doublest.c +++ binutils-gdb/gdb/doublest.c @@ -594,8 +594,13 @@ floatformat_classify (const struct float return float_normal; } - if (exponent == 0 && !mant_zero) - return float_subnormal; + if (exponent == 0) + { + if (mant_zero) + return float_zero; + else + return float_subnormal; + } if (exponent == fmt->exp_nan) { @@ -605,9 +610,6 @@ floatformat_classify (const struct float return float_nan; } - if (mant_zero) - return float_zero; - return float_normal; } Index: binutils-gdb/gdb/gdbtypes.c =================================================================== --- binutils-gdb.orig/gdb/gdbtypes.c +++ binutils-gdb/gdb/gdbtypes.c @@ -2935,6 +2935,16 @@ is_integral_type (struct type *t) || (TYPE_CODE (t) == TYPE_CODE_BOOL))); } +int +is_floating_type (struct type *t) +{ + t = check_typedef (t); + return + ((t != NULL) + && ((TYPE_CODE (t) == TYPE_CODE_FLT) + || (TYPE_CODE (t) == TYPE_CODE_DECFLOAT))); +} + /* Return true if TYPE is scalar. */ int Index: binutils-gdb/gdb/gdbtypes.h =================================================================== --- binutils-gdb.orig/gdb/gdbtypes.h +++ binutils-gdb/gdb/gdbtypes.h @@ -1927,6 +1927,8 @@ extern int can_dereference (struct type extern int is_integral_type (struct type *); +extern int is_floating_type (struct type *); + extern int is_scalar_type (struct type *type); extern int is_scalar_type_recursive (struct type *); Index: binutils-gdb/gdb/value.c =================================================================== --- binutils-gdb.orig/gdb/value.c +++ binutils-gdb/gdb/value.c @@ -29,10 +29,10 @@ #include "language.h" #include "demangle.h" #include "doublest.h" -#include "floatformat.h" #include "regcache.h" #include "block.h" #include "dfp.h" +#include "target-float.h" #include "objfiles.h" #include "valprint.h" #include "cli/cli-decode.h" @@ -2967,24 +2967,7 @@ unpack_double (struct type *type, const nosign = TYPE_UNSIGNED (type); if (code == TYPE_CODE_FLT) { - /* NOTE: cagney/2002-02-19: There was a test here to see if the - floating-point value was valid (using the macro - INVALID_FLOAT). That test/macro have been removed. - - It turns out that only the VAX defined this macro and then - only in a non-portable way. Fixing the portability problem - wouldn't help since the VAX floating-point code is also badly - bit-rotten. The target needs to add definitions for the - methods gdbarch_float_format and gdbarch_double_format - these - exactly describe the target floating-point format. The - problem here is that the corresponding floatformat_vax_f and - floatformat_vax_d values these methods should be set to are - also not defined either. Oops! - - Hopefully someone will add both the missing floatformat - definitions and the new cases for floatformat_is_valid (). */ - - if (!floatformat_is_valid (floatformat_from_type (type), valaddr)) + if (!target_float_is_valid (valaddr, type)) { *invp = 1; return 0.0; @@ -3027,6 +3010,21 @@ unpack_pointer (struct type *type, const return unpack_long (type, valaddr); } +bool +is_floating_value (struct value *val) +{ + struct type *type = check_typedef (value_type (val)); + + if (is_floating_type (type)) + { + if (!target_float_is_valid (value_contents (val), type)) + error (_("Invalid floating value found in program.")); + return true; + } + + return false; +} + /* Get the value of the FIELDNO'th field (which must be static) of TYPE. */ Index: binutils-gdb/gdb/value.h =================================================================== --- binutils-gdb.orig/gdb/value.h +++ binutils-gdb/gdb/value.h @@ -608,6 +608,10 @@ extern int print_address_demangle (const struct gdbarch *, CORE_ADDR, struct ui_file *, int); +/* Returns true if VAL is of floating-point type. In addition, + throws an error if the value is an invalid floating-point value. */ +extern bool is_floating_value (struct value *val); + extern LONGEST value_as_long (struct value *val); extern DOUBLEST value_as_double (struct value *val); extern CORE_ADDR value_as_address (struct value *val); Index: binutils-gdb/gdb/valarith.c =================================================================== --- binutils-gdb.orig/gdb/valarith.c +++ binutils-gdb/gdb/valarith.c @@ -26,6 +26,7 @@ #include "language.h" #include "doublest.h" #include "dfp.h" +#include "target-float.h" #include #include "infcall.h" @@ -1517,11 +1518,8 @@ value_logical_not (struct value *arg1) arg1 = coerce_array (arg1); type1 = check_typedef (value_type (arg1)); - if (TYPE_CODE (type1) == TYPE_CODE_FLT) - return 0 == value_as_double (arg1); - else if (TYPE_CODE (type1) == TYPE_CODE_DECFLOAT) - return decimal_is_zero (value_contents (arg1), TYPE_LENGTH (type1), - gdbarch_byte_order (get_type_arch (type1))); + if (is_floating_value (arg1)) + return target_float_is_zero (value_contents (arg1), type1); len = TYPE_LENGTH (type1); p = value_contents (arg1); Index: binutils-gdb/gdb/python/py-value.c =================================================================== --- binutils-gdb.orig/gdb/python/py-value.c +++ binutils-gdb/gdb/python/py-value.c @@ -21,7 +21,7 @@ #include "charset.h" #include "value.h" #include "language.h" -#include "dfp.h" +#include "target-float.h" #include "valprint.h" #include "infcall.h" #include "expression.h" @@ -1317,12 +1317,9 @@ valpy_nonzero (PyObject *self) if (is_integral_type (type) || TYPE_CODE (type) == TYPE_CODE_PTR) nonzero = !!value_as_long (self_value->value); - else if (TYPE_CODE (type) == TYPE_CODE_FLT) - nonzero = value_as_double (self_value->value) != 0; - else if (TYPE_CODE (type) == TYPE_CODE_DECFLOAT) - nonzero = !decimal_is_zero (value_contents (self_value->value), - TYPE_LENGTH (type), - gdbarch_byte_order (get_type_arch (type))); + else if (is_floating_value (self_value->value)) + nonzero = !target_float_is_zero (value_contents (self_value->value), + type); else /* All other values are True. */ nonzero = 1;