Re: [BUG] perf: does not compile on ARM because of Intel PT code

From: Stephane Eranian
Date: Tue Sep 01 2015 - 04:29:58 EST


On Tue, Sep 1, 2015 at 1:26 AM, Adrian Hunter <adrian.hunter@xxxxxxxxx> wrote:
> On 01/09/15 11:26, Stephane Eranian wrote:
>> Hi,
>>
>> If you compile tip.git on ARM32 (or ARM64, I bet) you compile link time errors:
>>
>> libperf.a(libperf-in.o): In function `intel_pt_synth_branch_sample':
>> /home/ubuntu/perfmon/linux/tools/perf/util/intel-pt.c:899: undefined
>> reference to `tsc_to_perf_time'
>> libperf.a(libperf-in.o): In function `intel_pt_synth_transaction_sample':
>> /home/ubuntu/perfmon/linux/tools/perf/util/intel-pt.c:992: undefined
>> reference to `tsc_to_perf_time'
>> libperf.a(libperf-in.o): In function `intel_pt_synth_instruction_sample':
>> /home/ubuntu/perfmon/linux/tools/perf/util/intel-pt.c:943: undefined
>> reference to `tsc_to_perf_time'
>> libperf.a(libperf-in.o): In function `intel_pt_process_event':
>> /home/ubuntu/perfmon/linux/tools/perf/util/intel-pt.c:1484: undefined
>> reference to `perf_time_to_tsc'
>> libperf.a(libperf-in.o): In function `intel_pt_process_itrace_start':
>> /home/ubuntu/perfmon/linux/tools/perf/util/intel-pt.c:1455: undefined
>> reference to `perf_time_to_tsc'
>> libperf.a(libperf-in.o): In function `intel_pt_process_switch':
>> /home/ubuntu/perfmon/linux/tools/perf/util/intel-pt.c:1405: undefined
>> reference to `perf_time_to_tsc'
>> /home/ubuntu/perfmon/linux/tools/perf/util/intel-pt.c:1427: undefined
>> reference to `perf_time_to_tsc'
>> libperf.a(libperf-in.o): In function `intel_bts_process_event':
>> /home/ubuntu/perfmon/linux/tools/perf/util/intel-bts.c:614: undefined
>> reference to `perf_time_to_tsc'
>>
>> That's because if AUXTRACE is detected, it pulls in x86 specific stuff
>> like intel-pt. This should
>> be decoupled.
>
> Ingo just pulled the fix for that a moment ago. Look for "perf tools: Fix
> build on powerpc broken by pt/bts"
>
Ok, good. I was testing on Linus' tree not tip.git.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/