On 07/12/2011 09:19 AM, David Ahern wrote: > The perf_event_attr struct has 2 __u32's at the top and need to > be swapped individually. With this change I was able to analyze > a perf.data collected in a 32-bit PPC VM on an x86 system. I > tested both 32-bit and 64-bit binaries for the Intel analysis > side; both read the PPC perf.data file correctly.
Similarly with this one. It replaces a call to perf_header__getbuffer64 with what it does inline -- only doing the swap of the attr struct after the initial 2 u32's. I spent 2 days tracking this down and would hate to lose it waiting for Arnaldo's return. David > > Signed-off-by: David Ahern <[email protected]> > --- > tools/perf/util/header.c | 9 ++++++++- > 1 files changed, 8 insertions(+), 1 deletions(-) > > diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c > index afb0849..8044c6f 100644 > --- a/tools/perf/util/header.c > +++ b/tools/perf/util/header.c > @@ -877,9 +877,16 @@ int perf_session__read_header(struct perf_session > *session, int fd) > struct perf_evsel *evsel; > off_t tmp; > > - if (perf_header__getbuffer64(header, fd, &f_attr, > sizeof(f_attr))) > + if (readn(fd, &f_attr, sizeof(f_attr)) <= 0) > goto out_errno; > > + if (header->needs_swap) { > + f_attr.attr.type = bswap_32(f_attr.attr.type); > + f_attr.attr.size = bswap_32(f_attr.attr.size); > + mem_bswap_64(&f_attr.attr.config, > + sizeof(struct perf_event_attr) - 8); > + } > + > tmp = lseek(fd, 0, SEEK_CUR); > evsel = perf_evsel__new(&f_attr.attr, i); > -- To unsubscribe from this list: send the line "unsubscribe linux-perf-users" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html
