Thread (14 messages) 14 messages, 4 authors, 2009-12-09

Re: [PATCH 6/8] Use proper types for do_div

From: Mark Wielaard <hidden>
Date: 2009-12-09 15:56:37

Hi Anton,

On Sat, 2009-11-28 at 01:33 +0300, Anton Vorontsov wrote: 
do_div accepts unsigned 64-bit integer type for dividend, signed types
would cause do_div's typecheck fail:

stat-common.c: In function 'needed_space':
stat-common.c:50: error: comparison of distinct pointer types lacks a cast
...same errors in time.c and tapset-timers.cxx's generated code...

A fix for time.c is special, on ppc32 cycles_t is 32-bit, so technically
we don't need do_div, but since the whole _stp_gettimeofday_ns() operates
on 64-bit types we'd better be safe and use uint64_t for the math.

Signed-off-by: Anton Vorontsov <redacted>
---
 runtime/stat-common.c |    8 ++++----
 runtime/time.c        |    3 ++-
 tapset-timers.cxx     |    2 +-
 3 files changed, 7 insertions(+), 6 deletions(-)
I don't immediately see anything wrong with the code. But on x86_64 this
does cause some regression test failures in testsuite/systemtap.maps. In
particular the following tests now fail (they all pass with this one
patch reverted):

Running /home/mark/src/systemtap/testsuite/systemtap.maps/elision.exp ...
FAIL: elision-1
FAIL: elision0
FAIL: elision1
FAIL: elision2
FAIL: elision3
Running /home/mark/src/systemtap/testsuite/systemtap.maps/ix.exp ...
FAIL: systemtap.maps/ix.stp
Running /home/mark/src/systemtap/testsuite/systemtap.maps/linear_large_neg.exp ...
FAIL: systemtap.maps/linear_large_neg.stp
Running /home/mark/src/systemtap/testsuite/systemtap.maps/linear_overunder.exp ...
FAIL: systemtap.maps/linear_overunder.stp
Running /home/mark/src/systemtap/testsuite/systemtap.maps/linear_under.exp ...
FAIL: systemtap.maps/linear_under.stp
Running /home/mark/src/systemtap/testsuite/systemtap.maps/log.exp ...
FAIL: systemtap.maps/log.stp
Running /home/mark/src/systemtap/testsuite/systemtap.maps/log_edge.exp ...
FAIL: systemtap.maps/log_edge.stp

Could you have a look?

Thanks,

Mark
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help