[Bug runtime/24681] New: Systemtap regression on 5.2.0-rc4+

[Bug runtime/24681] New: Systemtap regression on 5.2.0-rc4+

[vedabharath12345 at gmail dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=24681

            Bug ID: 24681
           Summary: Systemtap regression on 5.2.0-rc4+
           Product: systemtap
           Version: unspecified
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: runtime
          Assignee: systemtap at sourceware dot org
          Reporter: vedabharath12345 at gmail dot com
  Target Milestone: ---

I am install systemtap from source and running the following script:

global stats

probe kernel.function("do_try_to_free_pages").return {
   order = @entry($sc->order)
   # 10 seconds
   time = cpu_clock_ns(cpu()) - @entry(cpu_clock_ns(cpu()))
   stats["do_try_to_free_pages", execname(), order] <<< time
}

probe kernel.function("try_to_compact_pages").return {
   order = @entry($order)
   # 10 seconds
   time = cpu_clock_ns(cpu()) - @entry(cpu_clock_ns(cpu()))
   stats["try_to_compact_pages", execname(), order] <<< time
}

probe end {
   foreach ([f, en, order+] in stats) {
       printf("%s:%s: order %d allocation latency - ",f,en,order)
       printf("max/avg/min/count: %d/%d/%d/%d\n", @max(stats[f,en,order]),
@avg(stats[f,en,order]), @min(stats[f,en,order]), @count(stats[f,en,order]))
       #print(@hist_log(stats[f,en,order]))
   }
}

I run this with the following command:
sudo stap -v file.stp

I get the following regression:
Pass 1: parsed user script and 480 library scripts using
252832virt/58044res/7164shr/51388data kb, in 430usr/20sys/442real ms.
Pass 2: analyzed script: 5 probes, 6 functions, 4 embeds, 9 globals using
441388virt/248420res/9004shr/239944data kb, in 1560usr/600sys/2178real ms.
Pass 3: translated to C into
"/tmp/stap67x2ww/stap_1cc3aea9d7e2ba2d749b5c2064fb5c34_8894_src.c" using
441388virt/248548res/9132shr/239944data kb, in 10usr/10sys/7real ms.
In file included from <command-line>:0:0:
././include/linux/compiler_types.h:217:1: fatal error:
/tmp/stap67x2ww/stapconf_6255ecd233c87a15fc7a36976eb58ca9_664.h: No such file
or directory
 #endif /* __LINUX_COMPILER_TYPES_H */
 ^
compilation terminated.
In file included from <command-line>:0:0:
././include/linux/compiler_types.h:217:1: fatal error:
/tmp/stap67x2ww/stapconf_6255ecd233c87a15fc7a36976eb58ca9_664.h: No such file
or directory
 #endif /* __LINUX_COMPILER_TYPES_H */
 ^
compilation terminated.
make[2]: *** [/tmp/stap67x2ww/stap_1cc3aea9d7e2ba2d749b5c2064fb5c34_8894_src.o]
Error 1
make[2]: *** Waiting for unfinished jobs....
make[2]: ***
[/tmp/stap67x2ww/stap_1cc3aea9d7e2ba2d749b5c2064fb5c34_8894_aux_0.o] Error 1
make[1]: *** [_module_/tmp/stap67x2ww] Error 2
make: *** [sub-make] Error 2
WARNING: kbuild exited with status: 2
Pass 4: compiled C into "stap_1cc3aea9d7e2ba2d749b5c2064fb5c34_8894.ko" in
320usr/250sys/534real ms.
Pass 4: compilation failed.  [man error::pass4]

This is on Oracle Linux 7.5 with Linux 5.2.0-rc4+

On Ubuntu 16.04 with linux 5.2.0-rc4+, I get the following regressions:
Pass 1: parsed user script and 476 library scripts using
119984virt/86916res/5680shr/81644data kb, in 560usr/60sys/626real ms.
Pass 2: analyzed script: 5 probes, 6 functions, 1 embed, 9 globals using
180372virt/148564res/6768shr/142032data kb, in 1760usr/80sys/1836real ms.
Pass 3: translated to C into
"/tmp/stapdPYzar/stap_8971bfbfebae8e7afd75f0d8461b9919_8269_src.c" using
180372virt/148820res/7024shr/142032data kb, in 40usr/0sys/46real ms.
In file included from ./include/linux/mmzone.h:8:0,
                 from ./include/linux/gfp.h:6,
                 from
/usr/local/share/systemtap/runtime/linux/runtime_defines.h:20,
                 from /usr/local/share/systemtap/runtime/runtime_defines.h:8,
                 from
/tmp/stapdPYzar/stap_8971bfbfebae8e7afd75f0d8461b9919_8269_src.c:11:
./include/linux/spinlock.h:60:24: fatal error: asm/mmiowb.h: No such file or
directory
compilation terminated.
scripts/Makefile.build:278: recipe for target
'/tmp/stapdPYzar/stap_8971bfbfebae8e7afd75f0d8461b9919_8269_src.o' failed
make[1]: *** [/tmp/stapdPYzar/stap_8971bfbfebae8e7afd75f0d8461b9919_8269_src.o]
Error 1
Makefile:1595: recipe for target '_module_/tmp/stapdPYzar' failed
make: *** [_module_/tmp/stapdPYzar] Error 2
WARNING: kbuild exited with status: 2
Pass 4: compiled C into "stap_8971bfbfebae8e7afd75f0d8461b9919_8269.ko" in
10360usr/3010sys/13317real ms.
Pass 4: compilation failed.  [man error::pass4]

stap version:
Systemtap translator/driver (version 4.2/0.165, commit
release-4.1-27-g2d77f531a5d1)
Copyright (C) 2005-2019 Red Hat, Inc. and others
This is free software; see the source for copying conditions.
tested kernel versions: 2.6.18 ... 5.1-rc2
enabled features: PYTHON2 PYTHON3 NLS READLINE

-- 
You are receiving this mail because:
You are the assignee for the bug.

[Bug runtime/24681] Systemtap regression on 5.2.0-rc4+

[vedabharath12345 at gmail dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=24681

Bharath <vedabharath12345 at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |vedabharath12345 at gmail dot com

--- Comment #1 from Bharath <vedabharath12345 at gmail dot com> ---
Hey, On Oracle Linux 7.5

-- 
You are receiving this mail because:
You are the assignee for the bug.

[Bug runtime/24681] Systemtap regression on 5.2.0-rc4+

[vedabharath12345 at gmail dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=24681

--- Comment #2 from Bharath <vedabharath12345 at gmail dot com> ---
Hey, On Oracle Linux 7.5 with linux 5.2.0-rc4+ and with stap 4.2, I get the
following regression:

Pass 1: parsed user script and 476 library scripts using
215736virt/85988res/4624shr/81904data kb, in 330usr/20sys/360real ms.
Pass 2: analyzed script: 5 probes, 6 functions, 1 embed, 9 globals using
267040virt/138920res/6180shr/133208data kb, in 660usr/20sys/672real ms.
Pass 3: translated to C into
"/tmp/stapHe4lxX/stap_a1937e59a38430d00dc862c2e665d999_8306_src.c" using
267040virt/139048res/6308shr/133208data kb, in 0usr/0sys/7real ms.
In file included from /usr/local/share/systemtap/runtime/runtime.h:26:0,
                 from
/tmp/stapHe4lxX/stap_a1937e59a38430d00dc862c2e665d999_8306_src.c:26:
/usr/local/share/systemtap/runtime/linux/runtime.h:72:17: error:
‘udelay_simple’ undeclared here (not in a function)
 typedef typeof(&udelay_simple) udelay_simple_fn;
                 ^
/usr/local/share/systemtap/runtime/linux/probe_lock.h: In function
‘stp_lock_probe’:
/usr/local/share/systemtap/runtime/linux/runtime.h:73:23: error: cast from
pointer to integer of different size [-Werror=pointer-to-int-cast]
 #define udelay(x) ((* (udelay_simple_fn)(kallsyms_udelay_simple))((x)))
                       ^
/usr/local/share/systemtap/runtime/linux/probe_lock.h:58:5: note: in expansion
of macro ‘udelay’
     udelay (TRYLOCKDELAY);
     ^
/usr/local/share/systemtap/runtime/linux/runtime.h:73:21: error: invalid type
argument of unary ‘*’ (have ‘int’)
 #define udelay(x) ((* (udelay_simple_fn)(kallsyms_udelay_simple))((x)))
                     ^
/usr/local/share/systemtap/runtime/linux/probe_lock.h:58:5: note: in expansion
of macro ‘udelay’
     udelay (TRYLOCKDELAY);
     ^
/usr/local/share/systemtap/runtime/linux/runtime.h:73:23: error: cast from
pointer to integer of different size [-Werror=pointer-to-int-cast]
 #define udelay(x) ((* (udelay_simple_fn)(kallsyms_udelay_simple))((x)))
                       ^
/usr/local/share/systemtap/runtime/linux/probe_lock.h:69:5: note: in expansion
of macro ‘udelay’
     udelay (TRYLOCKDELAY);
     ^
/usr/local/share/systemtap/runtime/linux/runtime.h:73:21: error: invalid type
argument of unary ‘*’ (have ‘int’)
 #define udelay(x) ((* (udelay_simple_fn)(kallsyms_udelay_simple))((x)))
                     ^
/usr/local/share/systemtap/runtime/linux/probe_lock.h:69:5: note: in expansion
of macro ‘udelay’
     udelay (TRYLOCKDELAY);
     ^
/tmp/stapHe4lxX/stap_a1937e59a38430d00dc862c2e665d999_8306_src.c: At top level:
cc1: error: unrecognized command line option "-Wno-tautological-compare"
[-Werror]
cc1: all warnings being treated as errors
make[2]: *** [/tmp/stapHe4lxX/stap_a1937e59a38430d00dc862c2e665d999_8306_src.o]
Error 1
make[1]: *** [_module_/tmp/stapHe4lxX] Error 2
make: *** [sub-make] Error 2
WARNING: kbuild exited with status: 2
Pass 4: compiled C into "stap_a1937e59a38430d00dc862c2e665d999_8306.ko" in
8480usr/1310sys/9931real ms.
Pass 4: compilation failed.  [man error::pass4]

The previous error on Oracle Linux 7.5 with 5.2.0-rc4+ was with stap 3.2

-- 
You are receiving this mail because:
You are the assignee for the bug.

[Bug runtime/24681] Systemtap regression on 5.2.0-rc4+

[fche at redhat dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=24681

Frank Ch. Eigler <fche at redhat dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |fche at redhat dot com

--- Comment #3 from Frank Ch. Eigler <fche at redhat dot com> ---
The sample script above works for me on the 5.2.7-200.fc30.x86_64 kernel, with
both
Systemtap translator/driver (version 4.1/0.177/0.176, rpm 4.1-2.fc30) and
Systemtap translator/driver (version 4.2/0.177, commit
release-4.1-93-g5c6bda7fb779)
with a modern compiler, so I don't think there's a runtime<->kernel
incompatibility at play.

The no-tautological-compares warning/error needs to be worked around at the the
buildrun.cxx compiler-flag autoconf level.  If you'd like to give that part a
try, I can give some pointers.

-- 
You are receiving this mail because:
You are the assignee for the bug.

[Bug runtime/24681] Systemtap regression on 5.2.0-rc4+

[fche at redhat dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=24681

Frank Ch. Eigler <fche at redhat dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|---                         |OBSOLETE
             Status|UNCONFIRMED                 |RESOLVED

--- Comment #4 from Frank Ch. Eigler <fche at redhat dot com> ---
no recent report

-- 
You are receiving this mail because:
You are the assignee for the bug.