Screenshot instructions:
Windows
Mac
Red Hat Linux
Ubuntu
Click URL instructions:
Right-click on ad, choose "Copy Link", then paste here →
(This may not be possible with some types of ads)
|
From: Maynard Johnson <maynardj@us...> - 2010-10-13 16:33:17
|
On 10/13/2010 10:21 AM, Suravee Suthikulpanit wrote: > Hi, > > I have committed the patch. However, I run into two unrelated errors: > > 1. The compiler warned about possible uninitialized variable > > cc1plus: warnings being treated as errors > opimport.cpp: In function ‘void import_from_abi(const abi&, > const void*, size_t, odb_t*)’: > opimport.cpp:154: error: ‘key’ may be used uninitialized in this > function > make[2]: *** [opimport.o] Error 1 > make[2]: *** Waiting for unfinished jobs.... > mv -f .deps/abi.Tpo .deps/abi.Po > make[2]: Leaving directory > `/sandbox/oprofile-cvs/oprofile-test-clean/libabi' > make[1]: *** [install-recursive] Error 1 > make[1]: Leaving directory > `/sandbox/oprofile-cvs/oprofile-test-clean/libabi' > make: *** [install-recursive] Error 1 Suravee, I had posted a fix for this problem on April 8 and asked Will to review it. But then he and I both proceeded to forget about it. :-/ Little things fall through the cracks sometimes. Please review/test the patch I posted. If you ack it, I'll commit it. > > 2. make distcheck failed for the following reason: > > make check-TESTS > make[4]: Entering directory > `/sandbox/oprofile-cvs/oprofile-test-clean/oprofile-0.9.7cvs/_build/libop/tests' > PASS: cpu_type_tests > PASS: parse_event_tests > oprofile: could not open unit mask description > file ../../../events/mips/74K/unit_masks > FAIL: load_events_files_tests > PASS: alloc_counter_tests > PASS: mangle_tests > =================== > 1 of 5 tests failed > =================== > make[4]: *** [check-TESTS] Error 1 > make[4]: Leaving directory > `/sandbox/oprofile-cvs/oprofile-test-clean/oprofile-0.9.7cvs/_build/libop/tests' > make[3]: *** [check-am] Error 2 > make[3]: Leaving directory > `/sandbox/oprofile-cvs/oprofile-test-clean/oprofile-0.9.7cvs/_build/libop/tests' > make[2]: *** [check-recursive] Error 1 > make[2]: Leaving directory > `/sandbox/oprofile-cvs/oprofile-test-clean/oprofile-0.9.7cvs/_build/libop' > make[1]: *** [check-recursive] Error 1 > make[1]: Leaving directory > `/sandbox/oprofile-cvs/oprofile-test-clean/oprofile-0.9.7cvs/_build' > make: *** [distcheck] Error 1 > > For the 2nd case, I have investigated further and found that the > "mips/74K/[events,unitmask] and "mips/1004K/[events,unitmask]" are not > in the events/Makefile.am. Therefore, it is not installed in > the /usr/local/share/oprofile/mips. Whoops. My fault. I should have caught that when I reviewed the patch in the first place. I'll fix it. -Maynard > > Suravee > > > > On Thu, 2010-10-07 at 17:25 -0500, Maynard Johnson wrote: >> William Cohen wrote: >>> I was testing things out for the i386/arch_perfmon events and some of the results didn't look right. When reviewing "Table 30-1. UMask and Event Select Encodings for Pre-Defined Architectural Performance Events" in Intel® 64 and IA-32 Architectures Software Developer’s Manual Volume 3B: System Programming Guide, Part 2 , Order Number: 253669-034US March 2010 there seems to be some disagreement between events/arch/i386/arch_perfmon/events and Table 30-1. A similar "Table 2-13. Performance Monitor Features" in "Intel® Processor Identification and the CPUID Instruction Application Note 485" August 2009, doc no 241618-036 matches the IA-32 Manual and also disagrees with events/arch/i386/arch_perfmon/events. >>> >> >> Suravee, please review and commit if you approve. >> >> Thanks. >> -Maynard >> >>> On the particular machine I am testing things it out it appears that the BR_MISS_PRED_RETIRED is not available. It that the case that some processors do not have that enabled? Below is the details from /proc/cpuinfo: >>> >>> vendor_id : GenuineIntel >>> cpu family : 6 >>> model : 26 >>> model name : Intel(R) Xeon(R) CPU X5550 @ 2.67GHz >>> stepping : 4 >>> >>> Attached is a proposed patch that matches the oprofile values to the tables in the documents mentioned earlier and below is a ChangeLog entry. >>> >>> >>> 2010-10-07 William Cohen<wcohen@...> >>> >>> * events/i386/arch_perfmon/events: Correct filter values. >>> >>> >>> -Will >>> >>> Signed-off-by: William Cohen<wcohen@...> >>> >>> >>> >>> >>> >>> ------------------------------------------------------------------------------ >>> Beautiful is writing same markup. Internet Explorer 9 supports >>> standards for HTML5, CSS3, SVG 1.1, ECMAScript5, and DOM L2& L3. >>> Spend less time writing and rewriting code and more time creating great >>> experiences on the web. Be a part of the beta today. >>> http://p.sf.net/sfu/beautyoftheweb >>> >>> >>> >>> _______________________________________________ >>> oprofile-list mailing list >>> oprofile-list@... >>> https://lists.sourceforge.net/lists/listinfo/oprofile-list >> >> > > > |