Return-Path: Received: from sfi-mx-3.v28.ch3.sourceforge.com ([172.29.28.123] helo=mx.sourceforge.net) by sfs-ml-4.v29.ch3.sourceforge.com with esmtp (Exim 4.69) (envelope-from ) id 1NDx38-0001FE-31 for ltp-list@lists.sourceforge.net; Fri, 27 Nov 2009 09:23:34 +0000 Received: from smtp-gate3.ryobi.co.jp ([210.163.224.134]) by sfi-mx-3.v28.ch3.sourceforge.com with smtp (Exim 4.69) id 1NDx2y-0000ii-2f for ltp-list@lists.sourceforge.net; Fri, 27 Nov 2009 09:23:34 +0000 Received: from 5109496 ([10.113.10.40]) by po.rsoa.ryobi.co.jp (Post.Office MTA v4.1.1 release 20090730 ID# 201-058U1200L1200S0V41J) with ESMTP id jp for ; Fri, 27 Nov 2009 18:23:11 +0900 From: "Mitani" Date: Fri, 27 Nov 2009 18:23:11 +0900 Message-ID: <000001ca6f43$3ec6ac20$bc540460$@co.jp> MIME-Version: 1.0 Content-Language: ja Subject: [LTP] build failure in ia64 List-Id: Linux Test Project General Discussions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Errors-To: ltp-list-bounces@lists.sourceforge.net To: ltp-list@lists.sourceforge.net Hi, I ran "make" with ltp-2009-11-26.tar.gz. Kernel release is 2.6.18-164.el5. Architecture is ia64. But it failed with following error message. "make autotools" and "./configure" has been done before make. I use "make 3.81", "autoconf 2.61", "automake 1.10.2", and "m4 1.4.7". ------------ make[4]: Entering directory `/home/mitani/LTP/ltp-2009-11-26/testcases/kernel/controllers/ freezer' gcc -g -O2 -g -O2 -fno-strict-aliasing -pipe -Wall -I/home/mitani/LTP/ltp-2009-11-26/test cases/kernel/include -I../../../../include -I../../../../include -L../../../../lib vfor k.c -lltp -o vfork In file included from ../../syscalls/ptrace/ptrace.h:14, from vfork.c:44: /usr/include/sys/ptrace.h:127: error: array type has incomplete element type In file included from ../../syscalls/ptrace/ptrace.h:22, from vfork.c:44: /usr/include/asm/ptrace.h:208: error: redefinition of 'struct FU_pt_all_user_regs' make[4]: *** [vfork] Error 1 make[4]: Leaving directory `/home/mitani/LTP/ltp-2009-11-26/testcases/kernel/controllers/f reezer' ------------ Similar failure continues from ltp-2009-11-17.tar.gz at the latest. "ltp-2009-11-27.tar.gz" (today's version) seems to fail by another new failure before these problems. Perhaps, "ltp-2009-11-27.tar.gz" will fail by similar failure after this "another new failure" solved, I think. So I will report about "ltp-2009-11-26.tar.gz". The problems are following two. (1) Line 127 of "/usr/include/sys/ptrace.h" declares "array type has incomplete element type" (2) Line 208 of "/usr/include/asm/ptrace.h" declares same structure as "/usr/include/sys/ptrace.h". The files related to these problems are as follows: - ${LTPROOT}/testcases/kernel/controllers/freezer/vfork.c - ${LTPROOT}/testcases/kernel/syscalls/ptrace/ptrace.h It's added from ltp-2009-11-17.tar.gz at the latest. I call this file ".../ptrace/ptrace.h" afterward. - /usr/include/sys/ptrace.h I call this file ".../sys/ptrace.h" afterward. - /usr/include/asm/ptrace.h I call this file ".../asm/ptrace.h" afterward. Follows show ".../ptrace/ptrace.h", ".../sys/ptrace.h" and ".../asm/ptrace.h". ".../ptrace/ptrace.h": ------------ /* Ordering matters. Do not move this code block below the #include's */ #ifdef __ia64__ /* what a pos */ # define ia64_fpreg FU_ia64_fpreg # define pt_all_user_regs FU_pt_all_user_regs #endif #ifdef HAVE_SYS_PTRACE_H # include #elif defined(HAVE_LINUX_PTRACE_H) # include #endif #ifdef HAVE_SYS_REG_H # include #endif #ifdef HAVE_ASM_PTRACE_H # include #endif #undef FU_ia64_fpreg #undef FU_pt_all_user_regs ------------ ".../sys/ptrace.h": ------------ struct pt_all_user_regs { unsigned long nat; unsigned long cr_iip; unsigned long cfm; unsigned long cr_ipsr; unsigned long pr; unsigned long gr[32]; unsigned long br[8]; unsigned long ar[128]; struct ia64_fpreg fr[128]; ............ L.127 }; ------------ ".../asm/ptrace.h": ------------ struct pt_all_user_regs { ............ L.208 unsigned long nat; unsigned long cr_iip; unsigned long cfm; unsigned long cr_ipsr; unsigned long pr; unsigned long gr[32]; unsigned long br[8]; unsigned long ar[128]; struct ia64_fpreg fr[128]; }; ------------ I tried to avoid above "make" failure as follows: Avoiding of (1) problem: "# define ia64_fpreg FU_ia64_fpreg" definition of ".../ptrace/ptrace.h" influences "struct ia64_fpreg fr[128];" of ".../sys/ptrace.h", I thought. I searched "ia64_fpreg" for all LTP sources, but this structure name was not found. I have no ideas why "FU_ia64_fpreg" was defined. So I commented out "# define ia64_fpreg FU_ia64_fpreg". After that, this problem did not occur. Avoiding of (2) problem: If ".../sys/ptrace.h" and ".../asm/ptrace.h" are included at the same time in ".../ptrace/ptrace.h", this probrem occures. Probably, "HAVE_SYS_PTRACE_H" and "HAVE_ASM_PTRACE_H" become effective at the same time. I referred to "${LTPROOT}/include/config.h" and "${LTPROOT}/config.log", and found both defined "1". In "ltp-full-20090930.tgz", "vfork.c" didn't include ".../asm/ptrace.h" but ".../sys/ptrace.h". So I commented out "# include " from ".../ptrace/ptrace.h". After that, this problem did not occur in "vfork.c". But after this, "${LTPROOT}/testcases/kernel/syscalls/ptrace/ptrace04.c" failed. In "ptrace04.c", struct "pt_regs" is used, and it's declared in ".../asm/ptrace.h". After examination, I understood that "vfork.c" and six source files from "${LTPROOT}/testcases/kernel/syscalls/ptrace01.c" to ".../ptrace06.c" include "../ptrace/ptrace.h", and they occur "redefinition" compile failure without (2) revision of "../ptrace/ptrace.h". May I have some questions? (1) Why was "FU_ia64_fpreg" defined? I deleted this definition to let "make" succeed. But I'm worried whether this revision has a problem. (2) I refer "${LTPROOT}/include/config.h" in my system. It was as follows: ------------ [...] /* Define to 1 if you have the header file. */ #define HAVE_ASM_PTRACE_H 1 [...] /* Define to 1 if you have the header file. */ #define HAVE_SYS_PTRACE_H 1 [...] ------------ In this state, "redefinition" failure such as this time will happen, I think. But I don't know what kind of revision should I do. Regards-- -Tomonori Mitani ------------------------------------------------------------------------------ Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day trial. Simplify your report design, integration and deployment - and focus on what you do best, core application coding. Discover what's new with Crystal Reports now. http://p.sf.net/sfu/bobj-july _______________________________________________ Ltp-list mailing list Ltp-list@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/ltp-list