From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mx0a-001b2d01.pphosted.com (mx0b-001b2d01.pphosted.com [148.163.158.5]) by sourceware.org (Postfix) with ESMTPS id 12D31385B51F for ; Mon, 20 Feb 2023 20:24:56 +0000 (GMT) DMARC-Filter: OpenDMARC Filter v1.4.2 sourceware.org 12D31385B51F Authentication-Results: sourceware.org; dmarc=none (p=none dis=none) header.from=us.ibm.com Authentication-Results: sourceware.org; spf=pass smtp.mailfrom=us.ibm.com Received: from pps.filterd (m0098416.ppops.net [127.0.0.1]) by mx0b-001b2d01.pphosted.com (8.17.1.19/8.17.1.19) with ESMTP id 31KH6Mjs023888; Mon, 20 Feb 2023 20:24:53 GMT DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=ibm.com; h=message-id : subject : from : to : cc : date : in-reply-to : references : content-type : content-transfer-encoding : mime-version; s=pp1; bh=Nph2+jf/VAH2O++lnuEP+Dyo3GXehemaKafhdy4Meaw=; b=dH4QGmClAh9S0J5AGdcgsoVeYsxgqvuWWA8GmSk5gknBJEm2ZAAOhOYO/wcCK3GMQVEq 8oi3g42Aqllyg1NHcofZ/lfIDhR346ZsROpYgIs6FSdwN0805wGb1O7bLTI2Ip2rkYAs u7CwFx5gpZT+aEwWIRb5Ix9biNqseJcI4EZHzsCtbgEK9jXTG2pU82uWOWatf4zDF7eW K2mFlnxIwf6ZF5FhGKVc4GI3tr+5NHA3prPrNH297Zjerjlg2rSMdkxvHKYW89vc/8c3 1oi/nzFHRqYXUloJDjXZkbx3WuSM6Nb9TSVdG9Azm3Vddqlq6mwcLgySCUv9qu2967cy bA== Received: from pps.reinject (localhost [127.0.0.1]) by mx0b-001b2d01.pphosted.com (PPS) with ESMTPS id 3nv2atcpb1-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=NOT); Mon, 20 Feb 2023 20:24:53 +0000 Received: from m0098416.ppops.net (m0098416.ppops.net [127.0.0.1]) by pps.reinject (8.17.1.5/8.17.1.5) with ESMTP id 31KKCNfq009919; Mon, 20 Feb 2023 20:24:52 GMT Received: from ppma01dal.us.ibm.com (83.d6.3fa9.ip4.static.sl-reverse.com [169.63.214.131]) by mx0b-001b2d01.pphosted.com (PPS) with ESMTPS id 3nv2atcpav-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=NOT); Mon, 20 Feb 2023 20:24:52 +0000 Received: from pps.filterd (ppma01dal.us.ibm.com [127.0.0.1]) by ppma01dal.us.ibm.com (8.17.1.19/8.17.1.19) with ESMTP id 31KIIqKT011386; Mon, 20 Feb 2023 20:24:51 GMT Received: from smtprelay03.dal12v.mail.ibm.com ([9.208.130.98]) by ppma01dal.us.ibm.com (PPS) with ESMTPS id 3ntpa7aq4a-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=NOT); Mon, 20 Feb 2023 20:24:51 +0000 Received: from smtpav06.dal12v.mail.ibm.com (smtpav06.dal12v.mail.ibm.com [10.241.53.105]) by smtprelay03.dal12v.mail.ibm.com (8.14.9/8.14.9/NCO v10.0) with ESMTP id 31KKOoCi6947494 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-GCM-SHA384 bits=256 verify=OK); Mon, 20 Feb 2023 20:24:50 GMT Received: from smtpav06.dal12v.mail.ibm.com (unknown [127.0.0.1]) by IMSVA (Postfix) with ESMTP id 997AC58059; Mon, 20 Feb 2023 20:24:50 +0000 (GMT) Received: from smtpav06.dal12v.mail.ibm.com (unknown [127.0.0.1]) by IMSVA (Postfix) with ESMTP id 18E8658043; Mon, 20 Feb 2023 20:24:50 +0000 (GMT) Received: from li-e362e14c-2378-11b2-a85c-87d605f3c641.ibm.com (unknown [9.163.93.208]) by smtpav06.dal12v.mail.ibm.com (Postfix) with ESMTP; Mon, 20 Feb 2023 20:24:50 +0000 (GMT) Message-ID: <46c2c756475ba5923d7eed97996632a08285dd42.camel@us.ibm.com> Subject: Re: [PATCH ] PowerPC: fix for gdb.reverse/finish-precsave.exp and gdb.reverse/finish-reverse.exp From: Carl Love To: Ulrich Weigand , "gdb-patches@sourceware.org" , Bruno Larsen , "tdevries@suse.de" , "pedro@palves.net" Cc: cel@us.ibm.com Date: Mon, 20 Feb 2023 12:24:49 -0800 In-Reply-To: <041f62e9f26fd4a536bc90c34f072985582e6237.camel@de.ibm.com> References: <071f24ecf9b3a2bbbe8fee7db77492eb55c5f3ff.camel@us.ibm.com> <1d9b21914354bef6a290ac30673741e722e11757.camel@de.ibm.com> <3e3c9c40f07ab01c79fe10915e76ffa187c42ad9.camel@us.ibm.com> <122f5d2d3db9ef1979b0f8da927d005f32bba82c.camel@us.ibm.com> <011768e8-2b76-f8ed-1174-fbaa020b15e7@redhat.com> <78b464a1-e32e-c3da-85e4-7bfc322cc29f@redhat.com> <7848e9858b54e33e399b871774ffc0b5058c1736.camel@us.ibm.com> <65d44121-65f7-a212-79ec-07ce53c15ecb@suse.de> <9fe94c0979cb40979b0dea7693a901c2d9f66164.camel@us.ibm.com> <59417813-eb4a-baf8-4e5d-e225d6732f71@suse.de> <7a494157-494f-6adf-d533-bf373b0f054f@redhat.com> <71aa635593df0677811afb85409aa190bcfa4f6a.camel@us.ibm.com> <15864a6b87b25c93e99a28149f23138267735f2a.camel@us.ibm.com> <041f62e9f26fd4a536bc90c34f072985582e6237.camel@de.ibm.com> Content-Type: text/plain; charset="UTF-8" X-Mailer: Evolution 3.28.5 (3.28.5-18.el8) X-TM-AS-GCONF: 00 X-Proofpoint-GUID: 9eTEHQzGEPml-fPQnv8pHzXLQvvRJkPr X-Proofpoint-ORIG-GUID: hUO5kR0pwo8GB4NzsKFJljzwqYcpzyE7 Content-Transfer-Encoding: 7bit X-Proofpoint-UnRewURL: 0 URL was un-rewritten MIME-Version: 1.0 X-Proofpoint-Virus-Version: vendor=baseguard engine=ICAP:2.0.219,Aquarius:18.0.930,Hydra:6.0.562,FMLib:17.11.170.22 definitions=2023-02-20_17,2023-02-20_02,2023-02-09_01 X-Proofpoint-Spam-Details: rule=outbound_notspam policy=outbound score=0 bulkscore=0 malwarescore=0 suspectscore=0 mlxlogscore=999 adultscore=0 phishscore=0 mlxscore=0 clxscore=1011 lowpriorityscore=0 impostorscore=0 spamscore=0 priorityscore=1501 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.12.0-2212070000 definitions=main-2302200185 X-Spam-Status: No, score=-11.9 required=5.0 tests=BAYES_00,DKIM_SIGNED,DKIM_VALID,DKIM_VALID_EF,GIT_PATCH_0,KAM_SHORT,RCVD_IN_MSPIKE_H2,SPF_HELO_NONE,SPF_NONE,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: Ulrich, Bruno, GDB maintainers: Per the comments from Ulrich, I have updated the patch to address the comments about the source code and the testcase. I updated the new library procedure so it can be called from gdb_step_until. So gdb_step_until is now just a variant of the new repeat_cmd_until proceedure. The redundant if test has been removed. It was added to fix a regression testing failure I saw on the gdb.btrace/tailcall.exp on X86. I went back and redid the testing, rebuilding everything from scratch. I am not able to reproduce the test failure without the redundant check. Not sure why I initially saw the regression failure at this point? Perhaps I hadn't re-enabled the /proc/sys/kernel/perf_event_paranoid value? It must be 2 or lower for the test to run. Otherwise the test fails. Anyway, the regression tests on X86 run without the redundant check are passing on my laptop. I resolved the issues with the testing on X86 to make sure it isn't missing failures when testing gdb.reverse/step-indirect-call-thunk.exp. As mentioned in the previous email, the test does not run on PowerPC since the gcc command line options "-mindirect-branch=thunk" and "- mfunction-return=thunk" are not supported on PowerPC. I disabled the test on PowerPC in the updated patch. Retested on X86 generation 5 and PowerPC with no regressions found. Hopefully this version of the patch is acceptable. Carl ----------------------------------------------------- PowerPC: fix for gdb.reverse/finish-precsave.exp and gdb.reverse/finish-reverse.exp PPC64 multiple entry points. When executing in reverse the function finish_backward sets the break point at the alternate entry point, known as the global entry point (GEP) in PowerPC. However if the forward execution enters via the normal entry point, known as the local entry point (LEP) on PowerPC, reverse execution never sees the break point at the alternate entry point of the function. Reverse execution continues until the next break point is encountered thus stopping at the wrong place. This patch adds a new address to struct execution_control_state to hold the address of the alternate entry point (GEP). The finish_backwards function is updated, if the stopping point is between the normal entry point (LEP) and the end of the function, a breakpoint is set at the normal entry point. If the stopping point is between the entry points, a breakpoint is set to the alternate entry point. The patch fixes the behavior of the reverse-finish command on PowerPC to match the behavior of the command on other platforms, specifically X86. The patch does not change the behavior of the command on X86. A new test is added to verify the reverse-finish command on PowerPC correctly stops at the instruction where the function call is made. A new test, finish-reverse-next.exp, is added to verify the correct functionality of the reverse-finish command. The reverse-finish comand is expected to stop on the instruction that jumps to the function. Procedure proc_step_until from test gdb.reverse/step-indirect-call-thunk.exp is moved to lib/gdb.exp and renamed repeat_cmd_until. The patch fixes 11 regression errors in test gdb.reverse/finish-precsave.exp and 11 regression errors in test gdb.reverse/finish-reverse.exp. The patch has been tested on Power 10 and X86 processor with no new regression failures. --- gdb/infcmd.c | 47 ++-- gdb/infrun.c | 24 ++ .../gdb.reverse/finish-reverse-next.c | 91 +++++++ .../gdb.reverse/finish-reverse-next.exp | 224 ++++++++++++++++++ .../gdb.reverse/step-indirect-call-thunk.exp | 55 ++--- gdb/testsuite/lib/gdb.exp | 49 +++- 6 files changed, 421 insertions(+), 69 deletions(-) create mode 100644 gdb/testsuite/gdb.reverse/finish-reverse-next.c create mode 100644 gdb/testsuite/gdb.reverse/finish-reverse-next.exp diff --git a/gdb/infcmd.c b/gdb/infcmd.c index 77206fcbfe8..0fa5719d38b 100644 --- a/gdb/infcmd.c +++ b/gdb/infcmd.c @@ -1728,28 +1728,41 @@ finish_backward (struct finish_command_fsm *sm) no way that a function up the stack can have a return address that's equal to its entry point. */ - if (sal.pc != pc) - { - frame_info_ptr frame = get_selected_frame (nullptr); - struct gdbarch *gdbarch = get_frame_arch (frame); + CORE_ADDR alt_entry_point = sal.pc; + CORE_ADDR entry_point = alt_entry_point; + frame_info_ptr frame = get_selected_frame (nullptr); + struct gdbarch *gdbarch = get_frame_arch (frame); + + if (gdbarch_skip_entrypoint_p (gdbarch)) + /* Some architectures, like PowerPC use local and global entry points. + There is only one Entry Point (GEP = LEP) for other architectures. + The GEP is an alternate entry point. The LEP is the normal entry point. + The value of entry_point was initialized to the alternate entry point + (GEP). It will be adjusted to the normal entry point if the function + has two entry points. */ + entry_point = gdbarch_skip_entrypoint (gdbarch, sal.pc); - /* Set a step-resume at the function's entry point. Once that's - hit, we'll do one more step backwards. */ + if ((pc < alt_entry_point) || (pc > entry_point)) + { + /* We are in the body of the function. Set a breakpoint to backup to + the normal entry point. */ symtab_and_line sr_sal; - sr_sal.pc = sal.pc; + sr_sal.pc = entry_point; sr_sal.pspace = get_frame_program_space (frame); - insert_step_resume_breakpoint_at_sal (gdbarch, - sr_sal, null_frame_id); - - proceed ((CORE_ADDR) -1, GDB_SIGNAL_DEFAULT); + insert_step_resume_breakpoint_at_sal (gdbarch, sr_sal, + null_frame_id); } + else - { - /* We're almost there -- we just need to back up by one more - single-step. */ - tp->control.step_range_start = tp->control.step_range_end = 1; - proceed ((CORE_ADDR) -1, GDB_SIGNAL_DEFAULT); - } + /* We are either at one of the entry points or between the entry points. + If we are not at the alt_entry point, go back to the alt_entry_point + If we at the normal entry point step back one instruction, when we + stop we will determine if we entered via the entry point or the + alternate entry point. If we are at the alternate entry point, + single step back to the function call. */ + tp->control.step_range_start = tp->control.step_range_end = 1; + + proceed ((CORE_ADDR) -1, GDB_SIGNAL_DEFAULT); } /* finish_forward -- helper function for finish_command. FRAME is the diff --git a/gdb/infrun.c b/gdb/infrun.c index 87ab73c47a4..987dbd16ea4 100644 --- a/gdb/infrun.c +++ b/gdb/infrun.c @@ -1868,6 +1868,7 @@ struct execution_control_state struct target_waitstatus ws; int stop_func_filled_in = 0; + CORE_ADDR stop_func_alt_start = 0; CORE_ADDR stop_func_start = 0; CORE_ADDR stop_func_end = 0; const char *stop_func_name = nullptr; @@ -4680,6 +4681,11 @@ fill_in_stop_func (struct gdbarch *gdbarch, ecs->stop_func_start += gdbarch_deprecated_function_start_offset (gdbarch); + /* PowerPC functions have a Local Entry Point (LEP) and a Global + Entry Point (GEP). There is only one Entry Point (GEP = LEP) for + other architectures. */ + ecs->stop_func_alt_start = ecs->stop_func_start; + if (gdbarch_skip_entrypoint_p (gdbarch)) ecs->stop_func_start = gdbarch_skip_entrypoint (gdbarch, ecs->stop_func_start); @@ -7269,6 +7275,24 @@ process_event_stop_test (struct execution_control_state *ecs) } } + if (execution_direction == EXEC_REVERSE + && ecs->event_thread->control.proceed_to_finish + && ecs->event_thread->stop_pc () >= ecs->stop_func_alt_start + && ecs->event_thread->stop_pc () < ecs->stop_func_start) + { + /* We are executing the reverse-finish command. + If the system supports multiple entry points and we are finishing a + function in reverse. If we are between the entry points singe-step + back to the alternate entry point. If we are at the alternate entry + point -- just need to back up by one more single-step, which + should take us back to the function call. */ + ecs->event_thread->control.step_range_start + = ecs->event_thread->control.step_range_end = 1; + keep_going (ecs); + return; + + } + if (ecs->event_thread->control.step_range_end == 1) { /* It is stepi or nexti. We always want to stop stepping after diff --git a/gdb/testsuite/gdb.reverse/finish-reverse-next.c b/gdb/testsuite/gdb.reverse/finish-reverse-next.c new file mode 100644 index 00000000000..e95ee8e33a6 --- /dev/null +++ b/gdb/testsuite/gdb.reverse/finish-reverse-next.c @@ -0,0 +1,91 @@ +/* This testcase is part of GDB, the GNU debugger. + + Copyright 2012-2023 Free Software Foundation, Inc. + + 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 . */ + +/* The reverse finish command should return from a function and stop on + the first instruction of the source line where the function call is made. + Specifically, the behavior should match doing a reverse next from the + first instruction in the function. GDB should only require one reverse + step or next statement to reach the previous source code line. + + This test verifies the fix for gdb bugzilla: + + https://sourceware.org/bugzilla/show_bug.cgi?id=29927 + + PowerPC supports two entry points to a function. The normal entry point + is called the local entry point (LEP). The alternate entry point is called + the global entry point (GEP). The GEP is only used if the table of + contents (TOC) value stored in register r2 needs to be setup prior to + execution starting at the LEP. A function call via a function pointer + will entry via the GEP. A normal function call will enter via the LEP. + + This test has been expanded to include tests to verify the reverse-finish + command works properly if the function is called via the GEP. The original + test only verified the reverse-finish command for a normal call that used + the LEP. */ + +int +function2 (int a, int b) +{ + int ret = 0; + ret = ret + a + b; + return ret; +} + +int +function1 (int a, int b) // FUNCTION1 +{ + int ret = 0; + int (*funp) (int, int) = &function2; + /* The assembly code for this function when compiled for PowerPC is as + follows: + + 0000000010000758 : + 10000758: 02 10 40 3c lis r2,4098 <- GEP + 1000075c: 00 7f 42 38 addi r2,r2,32512 + 10000760: a6 02 08 7c mflr r0 <- LEP + 10000764: 10 00 01 f8 std r0,16(r1) + .... + + When the function is called on PowerPC with function1 (a, b) the call + enters at the Local Entry Point (LEP). When the function is called via + a function pointer, the Global Entry Point (GEP) for function1 is used. + The GEP sets up register 2 before reaching the LEP. + */ + ret = funp (a + 1, b + 2); + return ret; +} + +int +main(int argc, char* argv[]) +{ + int a, b; + int (*funp) (int, int) = &function1; + + /* Call function via Local Entry Point (LEP). */ + + a = 1; + b = 5; + + function1 (a, b); // CALL VIA LEP + + /* Call function via Global Entry Point (GEP). */ + a = 10; + b = 50; + + funp (a, b); // CALL VIA GEP + return 0; +} diff --git a/gdb/testsuite/gdb.reverse/finish-reverse-next.exp b/gdb/testsuite/gdb.reverse/finish-reverse-next.exp new file mode 100644 index 00000000000..1f53b649a7d --- /dev/null +++ b/gdb/testsuite/gdb.reverse/finish-reverse-next.exp @@ -0,0 +1,224 @@ +# Copyright 2008-2023 Free Software Foundation, Inc. + +# 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 . */ + +# This file is part of the GDB testsuite. It tests reverse stepping. +# Lots of code borrowed from "step-test.exp". + +# The reverse finish command should return from a function and stop on +# the first instruction of the source line where the function call is made. +# Specifically, the behavior should match doing a reverse next from the +# first instruction in the function. GDB should only take one reverse step +# or next statement to reach the previous source code line. + +# This testcase verifies the reverse-finish command stops at the first +# instruction in the source code line where the function was called. There +# are two scenarios that must be checked: +# 1) gdb is at the entry point instruction for the function +# 2) gdb is in the body of the function. + +# This test verifies the fix for gdb bugzilla: +# https://sourceware.org/bugzilla/show_bug.cgi?id=29927 + +# PowerPC supports two entry points to a function. The normal entry point +# is called the local entry point (LEP). The alternate entry point is called +# the global entry point (GEP). A function call via a function pointer +# will entry via the GEP. A normal function call will enter via the LEP. +# +# This test has been expanded to include tests to verify the reverse-finish +# command works properly if the function is called via the GEP. The original +# test only verified the reverse-finish command for a normal call that used +# the LEP. + +if ![supports_reverse] { + return +} + +standard_testfile + +if { [prepare_for_testing "failed to prepare" $testfile $srcfile] } { + return -1 +} + +runto_main +set target_remote [gdb_is_target_remote] + +if [supports_process_record] { + # Activate process record/replay. + gdb_test_no_output "record" "turn on process record for test1" +} + + +### TEST 1: reverse finish from the entry point instruction (LEP) in +### function1 when called using the normal entry point (LEP). + +# Set breakpoint at call to function1 in main. +set bp_LEP_test [gdb_get_line_number "CALL VIA LEP" $srcfile] +gdb_breakpoint $srcfile:$bp_LEP_test temporary + +# Continue to break point at function1 call in main. +gdb_continue_to_breakpoint \ + "stopped at function1 entry point instruction to stepi into function" \ + ".*$srcfile:$bp_LEP_test\r\n.*" + +# stepi until we see "{" indicating we entered function1 +repeat_cmd_until "stepi" "CALL VIA LEP" "{" "stepi into function1 call" "100" + +# The reverse-finish command should stop on the function call instruction +# which is the last instruction in the source code line. A reverse-next +# instruction should then stop at the first instruction in the same source +# code line. Another revers-next instruction stops at the previous source +# code line. +gdb_test "reverse-finish" ".*function1 \\(a, b\\); // CALL VIA LEP.*" \ + "reverse-finish function1 LEP call from LEP " +gdb_test "reverse-next" ".*function1 \\(a, b\\); // CALL VIA LEP" \ + "reverse next 1 LEP entry point function call from LEP" +gdb_test "reverse-next" ".*b = 5;.*" "reverse next 2, at b = 5, call from LEP" + + +gdb_test "reverse-continue" ".*" "setup for test 2" + +# Turn off record to clear logs and turn on again +gdb_test "record stop" "Process record is stopped.*" \ + "turn off process record for test1" +gdb_test_no_output "record" "turn on process record for test2" + + +### TEST 2: reverse finish from the body of function1. + +# Set breakpoint at call to function1 in main. +gdb_breakpoint $srcfile:$bp_LEP_test temporary + +# Continue to break point at function1 call in main. +gdb_continue_to_breakpoint \ + "at function1 entry point instruction to step to body of function" \ + ".*$srcfile:$bp_LEP_test\r\n.*" + +# do a step instruction to get to the body of the function +gdb_test "step" ".*int ret = 0;.*" "step test 1" + +# The reverse-finish command should stop on the function call instruction +# which is the last instruction in the source code line. A reverse-next +# instruction should then stop at the first instruction in the same source +# code line. Another revers-next instruction stops at the previous source +# code line. +gdb_test "reverse-finish" ".*function1 \\(a, b\\); // CALL VIA LEP.*" \ + "reverse-finish function1 LEP call from function body" +gdb_test "reverse-next" ".*function1 \\(a, b\\); // CALL VIA LEP.*" \ + "reverse next 1 LEP from function body" +gdb_test "reverse-next" ".*b = 5;.*" \ + "reverse next 2 at b = 5, from function body" + +gdb_test "reverse-continue" ".*" "setup for test 3" + +# Turn off record to clear logs and turn on again +gdb_test "record stop" "Process record is stopped.*" \ + "turn off process record for test2" +gdb_test_no_output "record" "turn on process record for test3" + + +### TEST 3: reverse finish from the alternate entry point instruction (GEP) in +### function1 when called using the alternate entry point (GEP). + +# Set breakpoint at call to funp in main. +set bp_GEP_test [gdb_get_line_number "CALL VIA GEP" $srcfile] +gdb_breakpoint $srcfile:$bp_GEP_test temporary + +# Continue to break point at funp call in main. +gdb_continue_to_breakpoint \ + "stopped at function1 entry point instruction to stepi into funp" \ + ".*$srcfile:$bp_GEP_test\r\n.*" + +# stepi until we see "{" indicating we entered function. +repeat_cmd_until "stepi" "CALL VIA GEP" "{" "stepi into funp call" + +# The reverse-finish command should stop on the function call instruction +# which is the last instruction in the source code line. A reverse-next +# instruction should then stop at the first instruction in the same source +# code line. Another revers-next instruction stops at the previous source +# code line. +gdb_test "reverse-finish" ".*funp \\(a, b\\);.*" \ + "function1 GEP call call from GEP" +gdb_test "reverse-next" ".*funp \\(a, b\\);.*" \ + "reverse next 1 GEP entry point function call from GEP" +gdb_test "reverse-next" ".*b = 50;.*" "reverse next 2 at b = 50, call from GEP" + +gdb_test "reverse-continue" ".*" "setup for test 4" + +# Turn off record to clear logs and turn on again +gdb_test "record stop" "Process record is stopped.*" \ + "turn off process record for test3" +gdb_test_no_output "record" "turn on process record for test4" + +### TEST 4: reverse finish from between the GEP and LEP in +### function1 when called using the alternate entry point (GEP). + +# Set breakpoint at call to funp in main. +set bp_GEP_test [gdb_get_line_number "CALL VIA GEP" $srcfile] +gdb_breakpoint $srcfile:$bp_GEP_test temporary + +# Continue to break point at funp call in main. +gdb_continue_to_breakpoint \ + "stopped at function1 entry point instruction to stepi into funp again" \ + ".*$srcfile:$bp_GEP_test\r\n.*" + +# stepi until we see "{" indicating we entered function. +repeat_cmd_until "stepi" "CALL VIA GEP" "{" "stepi into funp call again" + +# do one more stepi so we are between the GEP and LEP. +gdb_test "stepi" "{" "stepi to between GEP and LEP" + +# The reverse-finish command should stop on the function call instruction +# which is the last instruction in the source code line. A reverse-next +# instruction should then stop at the first instruction in the same source +# code line. Another revers-next instruction stops at the previous source +# code line. +gdb_test "reverse-finish" ".*funp \\(a, b\\);.*" \ + "function1 GEP call call from GEP again" +gdb_test "reverse-next" ".*funp \\(a, b\\);.*" \ + "reverse next 1 GEP entry point function call from GEP again" +gdb_test "reverse-next" ".*b = 50;.*" \ + "reverse next 2 at b = 50, call from GEP again" + +gdb_test "reverse-continue" ".*" "setup for test 5" + +# Turn off record to clear logs and turn on again +gdb_test "record stop" "Process record is stopped.*" \ + "turn off process record for test4" +gdb_test_no_output "record" "turn on process record for test5" + + +### TEST 5: reverse finish from the body of function 1 when calling using the +### alternate entrypoint (GEP). +gdb_breakpoint $srcfile:$bp_GEP_test temporary + +# Continue to break point at funp call. +gdb_continue_to_breakpoint \ + "at function1 entry point instruction to step to body of funp call" \ + ".*$srcfile:$bp_GEP_test\r\n.*" + +# Step into body of funp, called via GEP. +gdb_test "step" ".*int ret = 0;.*" "step test 2" + +# The reverse-finish command should stop on the function call instruction +# which is the last instruction in the source code line. A reverse-next +# instruction should then stop at the first instruction in the same source +# code line. Another revers-next instruction stops at the previous source +# code line. +gdb_test "reverse-finish" ".*funp \\(a, b\\);.*" \ + "reverse-finish function1 GEP call, from function body " +gdb_test "reverse-next" ".*funp \\(a, b\\);.*" \ + "reverse next 1 GEP call from function body" +gdb_test "reverse-next" ".*b = 50;.*" \ + "reverse next 2 at b = 50 from function body" diff --git a/gdb/testsuite/gdb.reverse/step-indirect-call-thunk.exp b/gdb/testsuite/gdb.reverse/step-indirect-call-thunk.exp index 94292d5eb9b..61fb4974b8e 100644 --- a/gdb/testsuite/gdb.reverse/step-indirect-call-thunk.exp +++ b/gdb/testsuite/gdb.reverse/step-indirect-call-thunk.exp @@ -17,6 +17,12 @@ require supports_reverse standard_testfile +if { [istarget "powerpc*-*-linux*"] } { + # GCC for PowerPC on linux does not support the -mindirect-branch and + # -mfunction-return command line options. + return 0 +} + set cflags {} lappend cflags debug lappend cflags additional_flags=-mindirect-branch=thunk @@ -36,39 +42,6 @@ if { ![runto_main] } { return -1 } -# Do repeated stepping COMMANDs in order to reach TARGET from CURRENT -# -# COMMAND is a stepping command -# CURRENT is a string matching the current location -# TARGET is a string matching the target location -# TEST is the test name -# -# The function issues repeated COMMANDs as long as the location matches -# CURRENT up to a maximum of 100 steps. -# -# TEST passes if the resulting location matches TARGET and fails -# otherwise. -# -proc step_until { command current target test } { - global gdb_prompt - - set count 0 - gdb_test_multiple "$command" "$test" { - -re "$current.*$gdb_prompt $" { - incr count - if { $count < 100 } { - send_gdb "$command\n" - exp_continue - } else { - fail "$test" - } - } - -re "$target.*$gdb_prompt $" { - pass "$test" - } - } -} - gdb_test_no_output "record" gdb_test "next" ".*" "record trace" @@ -88,20 +61,20 @@ gdb_test "reverse-next" "apply\.2.*" \ "reverse-step through thunks and over inc" # We can use instruction stepping to step into thunks. -step_until "stepi" "apply\.2" "indirect_thunk" "stepi into call thunk" -step_until "stepi" "indirect_thunk" "inc" \ +repeat_cmd_until "stepi" "apply\.2" "indirect_thunk" "stepi into call thunk" +repeat_cmd_until "stepi" "indirect_thunk" "inc" \ "stepi out of call thunk into inc" set alphanum_re "\[a-zA-Z0-9\]" set pic_thunk_re "__$alphanum_re*\\.get_pc_thunk\\.$alphanum_re* \\(\\)" -step_until "stepi" "(inc|$pic_thunk_re)" "return_thunk" "stepi into return thunk" -step_until "stepi" "return_thunk" "apply" \ +repeat_cmd_until "stepi" "(inc|$pic_thunk_re)" "return_thunk" "stepi into return thunk" +repeat_cmd_until "stepi" "return_thunk" "apply" \ "stepi out of return thunk back into apply" -step_until "reverse-stepi" "apply" "return_thunk" \ +repeat_cmd_until "reverse-stepi" "apply" "return_thunk" \ "reverse-stepi into return thunk" -step_until "reverse-stepi" "return_thunk" "inc" \ +repeat_cmd_until "reverse-stepi" "return_thunk" "inc" \ "reverse-stepi out of return thunk into inc" -step_until "reverse-stepi" "(inc|$pic_thunk_re)" "indirect_thunk" \ +repeat_cmd_until "reverse-stepi" "(inc|$pic_thunk_re)" "indirect_thunk" \ "reverse-stepi into call thunk" -step_until "reverse-stepi" "indirect_thunk" "apply" \ +repeat_cmd_until "reverse-stepi" "indirect_thunk" "apply" \ "reverse-stepi out of call thunk into apply" diff --git a/gdb/testsuite/lib/gdb.exp b/gdb/testsuite/lib/gdb.exp index faa0ac05a9a..b10555fe5fb 100644 --- a/gdb/testsuite/lib/gdb.exp +++ b/gdb/testsuite/lib/gdb.exp @@ -9267,31 +9267,58 @@ gdb_caching_proc arm_cc_for_target { # Step until the pattern REGEXP is found. Step at most # MAX_STEPS times, but stop stepping once REGEXP is found. -# +# START matches current location # If REGEXP is found then a single pass is emitted, otherwise, after # MAX_STEPS steps, a single fail is emitted. # # TEST_NAME is the name used in the pass/fail calls. -proc gdb_step_until { regexp {test_name ""} {max_steps 10} } { +proc gdb_step_until { regexp {test_name ""} {current ""} \ + { max_steps 10 } } { + if { $current == "" } { + set current "\}" + } + if { $test_name == "" } { + set test_name "stepping until regexp" + } + + repeat_cmd_until "step" $current $regexp $test_name "10" +} + +# Do repeated stepping COMMANDs in order to reach TARGET from CURRENT +# +# COMMAND is a stepping command +# CURRENT is a string matching the current location +# TARGET is a string matching the target location +# TEST is the test name +# MAX_STEPS is number of steps attempted before fail is emitted +# +# The function issues repeated COMMANDs as long as the location matches +# CURRENT up to a maximum of 100 steps. +# +# TEST passes if the resulting location matches TARGET and fails +# otherwise. +# +proc repeat_cmd_until { command current target test_name {max_steps 100} } { + global gdb_prompt if { $test_name == "" } { set test_name "stepping until regexp" } set count 0 - gdb_test_multiple "step" "$test_name" { - -re "$regexp\r\n$::gdb_prompt $" { - pass $test_name - } - -re ".*$::gdb_prompt $" { - if {$count < $max_steps} { - incr count - send_gdb "step\n" + gdb_test_multiple "$command" "$test_name" { + -re "$current.*$gdb_prompt $" { + incr count + if { $count < $max_steps } { + send_gdb "$command\n" exp_continue } else { - fail $test_name + fail "$test_name" } } + -re "$target.*$gdb_prompt $" { + pass "$test_name" + } } } -- 2.37.2