From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 26044 invoked by alias); 10 Nov 2010 07:30:33 -0000 Received: (qmail 25959 invoked by uid 22791); 10 Nov 2010 07:30:29 -0000 X-SWARE-Spam-Status: No, hits=-1.9 required=5.0 tests=AWL,BAYES_00,TW_LW,T_RP_MATCHES_RCVD X-Spam-Check-By: sourceware.org Received: from e24smtp03.br.ibm.com (HELO e24smtp03.br.ibm.com) (32.104.18.24) by sourceware.org (qpsmtpd/0.43rc1) with ESMTP; Wed, 10 Nov 2010 07:29:52 +0000 Received: from mailhub3.br.ibm.com (mailhub3.br.ibm.com [9.18.232.110]) by e24smtp03.br.ibm.com (8.14.4/8.13.1) with ESMTP id oAA7M3Bl007214 for ; Wed, 10 Nov 2010 05:22:03 -0200 Received: from d24av04.br.ibm.com (d24av04.br.ibm.com [9.8.31.97]) by mailhub3.br.ibm.com (8.13.8/8.13.8/NCO v10.0) with ESMTP id oAA7aVgg2596900 for ; Wed, 10 Nov 2010 05:36:31 -0200 Received: from d24av04.br.ibm.com (loopback [127.0.0.1]) by d24av04.br.ibm.com (8.14.4/8.13.1/NCO v10.0 AVout) with ESMTP id oAA7TjjE012881 for ; Wed, 10 Nov 2010 05:29:45 -0200 Received: from evol.ibm.com ([9.8.8.94]) by d24av04.br.ibm.com (8.14.4/8.13.1/NCO v10.0 AVin) with ESMTP id oAA7TjZj012851 for ; Wed, 10 Nov 2010 05:29:45 -0200 From: Edjunior Barbosa Machado To: gdb-patches@sourceware.org Subject: [PATCH] testsuite: asm-source.exp: add powerpc64 support Date: Wed, 10 Nov 2010 07:30:00 -0000 Message-Id: <1289374182-14133-1-git-send-email-emachado@linux.vnet.ibm.com> X-IsSubscribed: yes Mailing-List: contact gdb-patches-help@sourceware.org; run by ezmlm Precedence: bulk List-Id: List-Subscribe: List-Archive: List-Post: List-Help: , Sender: gdb-patches-owner@sourceware.org X-SW-Source: 2010-11/txt/msg00144.txt.bz2 Hi, this patch intents to add powerpc64 support to the asm-source.exp testcase. It contains a new include file (powerpc64.inc) which defines macros for ppc64 asm. As some architectures (like ppc64) can have one or more instructions after the subroutine call, the testcase was modified in order to expect the caller line again as well as the next line when issuing "finish" or "return" (as mentioned at http://sourceware.org/ml/gdb/2010-11/msg00032.html). It also has other minor changes, such as expect '.' prefixing ppc64's symbols and call 'sys_exit' sequence on gdbasm_exit0 macro for ppc32. Tested on x86, ppc32 and ppc64. Is that ok? Thanks, -- Edjunior gdb/testsuite/ 2010-11-10 Edjunior Machado * asm-source.exp: Add new target "powerpc64"; expect leading `.' on ppc64's symbols; "finish" and "return" commands must accept to show the caller line again as well as the line after. * powerpc64.inc: New file * powerpc.inc: Use 'sys_exit' --- gdb/testsuite/gdb.asm/asm-source.exp | 44 ++++++++++++++++++++++--- gdb/testsuite/gdb.asm/powerpc.inc | 6 ++-- gdb/testsuite/gdb.asm/powerpc64.inc | 57 ++++++++++++++++++++++++++++++++++ 3 files changed, 98 insertions(+), 9 deletions(-) create mode 100644 gdb/testsuite/gdb.asm/powerpc64.inc diff --git a/gdb/testsuite/gdb.asm/asm-source.exp b/gdb/testsuite/gdb.asm/asm-source.exp index ca342bd..9053711 100644 --- a/gdb/testsuite/gdb.asm/asm-source.exp +++ b/gdb/testsuite/gdb.asm/asm-source.exp @@ -101,7 +101,15 @@ switch -glob -- [istarget] { set asm-arch mips } "powerpc*-*" { - set asm-arch powerpc + if { [is_lp64_target] } { + set asm-arch powerpc64 + set asm-flags "-a64 -I${srcdir}/${subdir} -I${objdir}/${subdir}" + append link-flags " -m elf64ppc" + } else { + set asm-arch powerpc + set asm-flags "-a32 -I${srcdir}/${subdir} -I${objdir}/${subdir}" + append link-flags " -m elf32ppc" + } } "sh*-*-*" { set asm-arch sh @@ -362,8 +370,19 @@ gdb_test "info source" \ "info source asmsrc1.s" # Try 'finishing' from foo3 -gdb_test "finish" "Run till exit from.*\[\r\n\]$line_call_foo3_again\[ \t\]+gdbasm_call foo3" \ - "finish from foo3" +# Some architectures will have one or more instructions after the +# call instruction which still is part of the call sequence, so we +# must be prepared for a "finish" to show us the caller line +# again as well as the statement after. +gdb_test_multiple "finish" "finish from foo3" { + -re "Run till exit from.*\[\r\n\]$line_call_foo3\[ \t\]+gdbasm_call foo3.*$gdb_prompt $" { + pass "finish from foo3" + gdb_test "s" ".*" "s after finish" + } + -re "Run till exit from.*\[\r\n\]$line_call_foo3_again\[ \t\]+gdbasm_call foo3.*$gdb_prompt $" { + pass "finish from foo3" + } +} # Try 'info source' from asmsrc2.s gdb_test "info source" \ @@ -400,15 +419,28 @@ gdb_test_multiple "info sources" "info sources" { # Try 'info line' gdb_test "info line" \ - "Line $line_call_foo3_again of.*asmsrc2.s.*starts at.* and ends at.*." \ + "Line $line_call_foo3_again of.*asmsrc2.s.*starts at.*<\\.?foo2+.*> and ends at.*<\\.?foo2+.*>." \ "info line" # Try 'nexting' over next call to foo3 gdb_test "next" "$line_foo2_leave\[ \t\]+gdbasm_leave" "next over foo3" # Try 'return' from foo2 -gdb_test "return" "\#0 main .*$line_main_exit\[ \t\]+gdbasm_exit0" "return from foo2" \ - "Make (foo2|selected stack frame) return now\?.*" "y" +# Like "finish", "return" command also can return to the caller +# line again or the statement after, depending on the architecture. +gdb_test_multiple "return" "return from foo2" { + -re "Make (foo2|selected stack frame) return now\?.*" { + send_gdb "y\n" + exp_continue + } + -re "\#0.*main .*$line_call_foo2\[ \t\]+gdbasm_call foo2.*$gdb_prompt $" { + pass "return from foo2" + gdb_test "s" ".*" "s after return" + } + -re "\#0.*main .*$line_main_exit\[ \t\]+gdbasm_exit0.*$gdb_prompt $" { + pass "return from foo2" + } +} # Disassemble something, check the output proc test_dis { command var } { diff --git a/gdb/testsuite/gdb.asm/powerpc.inc b/gdb/testsuite/gdb.asm/powerpc.inc index a0797bf..ab700c2 100644 --- a/gdb/testsuite/gdb.asm/powerpc.inc +++ b/gdb/testsuite/gdb.asm/powerpc.inc @@ -29,9 +29,9 @@ comment "exit (0)" .macro gdbasm_exit0 - comment "Don't know how to exit, but this will certainly halt..." - li 0, 0 - lwz 0, 0(0) + comment "sys_exit" + li 0, 1 + sc .endm comment "crt0 startup" diff --git a/gdb/testsuite/gdb.asm/powerpc64.inc b/gdb/testsuite/gdb.asm/powerpc64.inc new file mode 100644 index 0000000..7265c07 --- /dev/null +++ b/gdb/testsuite/gdb.asm/powerpc64.inc @@ -0,0 +1,57 @@ + comment "subroutine declare" + .purgem gdbasm_declare + .macro gdbasm_declare name + .section ".opd","aw" +\name: + .quad .\name, .TOC.@tocbase, 0 + .section ".text" + .type \name, @function +.\name: + .endm + + comment "subroutine prologue" + .macro gdbasm_enter + stdu 1, -112(1) + mflr 0 + std 0, 128(1) + .endm + + comment "subroutine epilogue" + .macro gdbasm_leave + ld 0, 128(1) + mtlr 0 + ld 1, 0(1) + blr + .endm + + .macro gdbasm_call subr + bl \subr + nop + .endm + + .macro gdbasm_several_nops + nop + nop + nop + nop + .endm + + comment "exit (0)" + .macro gdbasm_exit0 + comment "sys_exit" + li 0, 1 + sc + .endm + + comment "crt0 startup" + .macro gdbasm_startup + .p2align 2 + .endm + + comment "Declare a data variable" + .purgem gdbasm_datavar + .macro gdbasm_datavar name value + .section ".data" +\name: + .long \value + .endm -- 1.7.1