XVilka
commented
6 years ago Closed XVilka closed 5 years ago
XVilka
commented
6 years ago 
Maijin
commented
6 years ago We will get more as we continue the translation to new
XVilka
commented
5 years ago In fact the count of failing tests increased from 55 to 159. Judging by the failing tests radare2 is hardly usable on Windows... log.txt
xarkes
commented
5 years ago We need to reach 0 and make appveyor become red when any test fail. I don't really have the time to do this, but anyone working on Windows could help on that.
a1ext
commented
5 years ago Some of the failed tests cases are related to /dev/null usage (not available on windows, nul has to be used instead) and environment variables like this one test:
%ENV=World
!echo $ENV
-World
+$ENVthrough subcommand, env var have to be accessed by echo %ENV%.
P.S. I don't see there a way to turn on/off some test cases depending on OS, so such test cases which differs in platform-related things can be handled properly by a test writer.
Maijin
commented
5 years ago Those cases must be rewritten to not use /dev/null nor nul. Not disabled.
XVilka
commented
5 years ago XVilka
commented
5 years ago @a1ext you can see that most of the failing tests are those using !rasm2 or !rahash2 calls. So by reusing internal r2 commands you can fix the tests.
a1ext
commented
5 years ago Lets fix them one-by-one.
Next one is the following:
[00:08:53] [XX] db\tools\rahash2 rahash2 -s hello\x31 992
[00:08:53] $ r2 -escr.utf8=0 -escr.color=0 -escr.interactive=0 -N -Q -i C:\Users\appveyor\AppData\Local\Temp\1\tmp-2356OLtWbSqcQN45.tmp -
[00:08:53] !rahash2 -qa md5 -s hello\\x31
[00:08:53]
[00:08:53] -203ad5ffa1d7c650ad681fdff3965cd2
[00:08:53] +827dd13a4b222a80edbc195239fd7c5c
[00:08:53]
[00:08:53] EXPECT=<<RUN
[00:08:53] 827dd13a4b222a80edbc195239fd7c5cIt caused by double-backslash \\ because on windows it \ is valid shell character so it hasn't be escaped. So, I dunno how to fix this
radare
commented
5 years ago Even in quotes? We can also feed an hexpair string
On 21 Mar 2019, at 09:26, a1ext notifications@github.com wrote:
Lets fix them one-by-one.
Next one is the following:
[00:08:53] [XX] db\tools\rahash2 rahash2 -s hello\x31 992 [00:08:53] $ r2 -escr.utf8=0 -escr.color=0 -escr.interactive=0 -N -Q -i C:\Users\appveyor\AppData\Local\Temp\1\tmp-2356OLtWbSqcQN45.tmp - [00:08:53] !rahash2 -qa md5 -s hello\x31 [00:08:53] [00:08:53] -203ad5ffa1d7c650ad681fdff3965cd2 [00:08:53] +827dd13a4b222a80edbc195239fd7c5c [00:08:53] [00:08:53] EXPECT=<<RUN [00:08:53] 827dd13a4b222a80edbc195239fd7c5c It caused by double-backslash \ because on windows it \ is valid shell character so it hasn't be escaped. So, I dunno how to fix this
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or mute the thread.
a1ext
commented
5 years ago yeah,
XVilka
commented
5 years ago Also something strange happens in demangling:
[XX] db\formats\mangling\mangling Abbreviation Of Repetead Names (Name with Template Arguments) demangler test 1285
$ r2 -escr.utf8=0 -escr.color=0 -escr.interactive=0 -N -Q -i C:\Users\appveyor\AppData\Local\Temp\1\tmp-4216W63kLQr4i94O.tmp -
"!rabin2 -D msvc ?xyz@?\$abc@V?\$def@H@@PAX@@YAXXZ"
-void __cdecl abc<class def<int>, void *>::xyz(void)
+?xyz@?\$abc@V?\$def@H@@PAX@@YAXXZ
EXPECT=<<EOF
?xyz@?\$abc@V?\$def@H@@PAX@@YAXXZ
[XX] db\formats\mangling\mangling Abbreviation Of Name with template demangler test 1372
$ r2 -escr.utf8=0 -escr.color=0 -escr.interactive=0 -N -Q -i C:\Users\appveyor\AppData\Local\Temp\1\tmp-4216XTjIkyr6272K.tmp -
"!rabin2 -D msvc ??\$MyTemplateFunction@VClass1@@@Class1@@QAEXPAV0@@Z"
-public: void __thiscall Class1::MyTemplateFunction<class Class1>(class Class1 *)
+??\$MyTemplateFunction@VClass1@@@Class1@@QAEXPAV0@@Z
EXPECT=<<EOF
??\$MyTemplateFunction@VClass1@@@Class1@@QAEXPAV0@@Z
[**] db\formats\mangling\mangling OK 1465 BR 86 XX 20 FX 15
[XX] db\formats\mangling\rust rust static type with hash 1224
$ r2 -escr.utf8=0 -escr.color=0 -escr.interactive=0 -N -Q -i C:\Users\appveyor\AppData\Local\Temp\1\tmp-4216wWX65sYmJpDp.tmp -
"!rabin2 -D rust _ZN96_\$LT\$core..fmt..Write..write_fmt..Adapter\$LT\$\$u27\$a\$C\$\$u20\$T\$GT\$\$u20\$as\$u20\$core..fmt..Write\$GT\$9write_str17he4f4768a2f446facE"
-<core::fmt::Write::write_fmt::Adapter<'a, T> as core::fmt::Write>::write_str::he4f4768a2f446fac
+_ZN96_\$LT\$core..fmt..Write..write_fmt..Adapter\$LT\$\$u27\$a\$C\$\$u20\$T\$GT\$\$u20\$as\$u20\$core..fmt..Write\$GT\$9write_str17he4f4768a2f446facE
EXPECT=<<EOF
_ZN96_\$LT\$core..fmt..Write..write_fmt..Adapter\$LT\$\$u27\$a\$C\$\$u20\$T\$GT\$\$u20\$as\$u20\$core..fmt..Write\$GT\$9write_str17he4f4768a2f446facE
[XX] db\formats\mangling\rust rust static type 1275
$ r2 -escr.utf8=0 -escr.color=0 -escr.interactive=0 -N -Q -i C:\Users\appveyor\AppData\Local\Temp\1\tmp-4216tWaLTOTFkhnb.tmp -
"!rabin2 -D rust _ZN71_\$LT\$Test\$u20\$\$u2b\$\$u20\$\$u27\$static\$u20\$as\$u20\$foo..Bar\$LT\$Test\$GT\$\$GT\$3barE"
-<Test + 'static as foo::Bar<Test>>::bar
+_ZN71_\$LT\$Test\$u20\$\$u2b\$\$u20\$\$u27\$static\$u20\$as\$u20\$foo..Bar\$LT\$Test\$GT\$\$GT\$3barE
EXPECT=<<EOFThis one is probably due to some format string difference between GCC and MSVC:
[XX] db\cmd\cmd_help "? -1" 408
$ r2 -escr.utf8=0 -escr.color=0 -escr.interactive=0 -N -Q -i C:\Users\appveyor\AppData\Local\Temp\1\tmp-42164UGuM31k6iLu.tmp --
? -1
int64 -1
hex 0xffffffffffffffff
octal 01777777777777777777777
-unit 16.0E
+unit 16E
segment fffff000:0fff
string "\xffAnd I remember @radare recently met these on another platform:
[**] db\cmd\cmd_help OK 2803 BR 142 XX 25 FX 19
[XX] db\cmd\cmd_i izzz* (file x86_64) 557
$ r2 -escr.utf8=0 -escr.color=0 -escr.interactive=0 -N -Q -i C:\Users\appveyor\AppData\Local\Temp\1\tmp-4216KERD4Q8IuEkK.tmp ../bins/elf/analysis/hello-linux-x86_64
izzz*~:0
izzz*~:1
f str.8 10 @ 0x00400034
-Cs 10 @ 0x00400034
+Cs 2314987372554 @ 0x00400034
EXPECT=<<EOF
f str.8 10 @ 0x00400034
Cs 2314987372554 @ 0x00400034
[XX] db\cmd\cmd_i utf8 (possibly malformed) filenames with non-printable chars 483
$ r2 -escr.utf8=0 -escr.color=0 -escr.interactive=0 -N -Q -i C:\Users\appveyor\AppData\Local\Temp\1\tmp-4216NITLz91cdKtM.tmp -
cd `e dir.tmp`
e io.cache=true
"(show_fname fname, wz $0, wtf `prz`, o `prz`, i~file, ij, rm `prz`)"
.(show_fname A\x1b¢\302\200€𝄞\363\240\201\201\\.bin)
.(show_fname B\x1b¢\302\200€𝄞\363\240\201\201\\.bin)
# INVALID FILENAME .(show_fname B\x1b¢\302\200\200€𝄞\363\240\201\201\\.bin)
-file A�€𝄞\.bin
-{"core":{"type":"","file":"A\u001b¢\u0080€𝄞\udb40\udc41\\.bin","fd":4,"size":256,"humansz":"256","iorw":true,"mode":"r-x","obsz":0,"block":256,"format":"any"}}
-file B�€𝄞\.bin
-{"core":{"type":"","file":"B\u001b¢\u0080€𝄞\udb40\udc41\\.bin","fd":5,"size":256,"humansz":"256","iorw":true,"mode":"r-x","obsz":0,"block":256,"format":"any"}}
+file malloc://512
+{"core":{"type":"","file":"malloc://512","fd":3,"size":512,"humansz":"512","iorw":true,"mode":"rwx","obsz":0,"block":256,"format":"any"}}
+file malloc://512
+{"core":{"type":"","file":"malloc://512","fd":3,"size":512,"humansz":"512","iorw":true,"mode":"rwx","obsz":0,"block":256,"format":"any"}}
EXPECT=<<EOF
file malloc://512
{"core":{"type":"","file":"malloc://512","fd":3,"size":512,"humansz":"512","iorw":true,"mode":"rwx","obsz":0,"block":256,"format":"any"}}
file malloc://512
{"core":{"type":"","file":"malloc://512","fd":3,"size":512,"humansz":"512","iorw":true,"mode":"rwx","obsz":0,"block":256,"format":"any"}}
[XX] db\cmd\cmd_i iS= 546
$ r2 -escr.utf8=0 -escr.color=0 -escr.interactive=0 -N -Q -i C:\Users\appveyor\AppData\Local\Temp\1\tmp-4216GYS48qWjbG4j.tmp ../bins/mach0/fatmach0-3true
iS=
-00* 0x100000ef8 |##--------| 0x100000f3f 71 r-x 0.__TEXT.__text
-01 0x100000f40 |--#-------| 0x100000f46 6 r-x 1.__TEXT.__symbol_stub1
-02 0x100000f46 |--#-------| 0x100000f60 26 r-x 2.__TEXT.__stub_helper
-03 0x100000f60 |--####----| 0x100000fb4 84 r-x 3.__TEXT.__unwind_info
-04 0x100000fb8 |-----###--| 0x100001000 72 r-x 4.__TEXT.__eh_frame
-05 0x100001000 |-------#--| 0x100001010 16 rw- 5.__DATA.__nl_symbol_ptr
-06 0x100001010 |-------#--| 0x100001018 8 rw- 6.__DATA.__la_symbol_ptr
-07 0x100001020 |--------##| 0x100001048 40 rw- 7.__DATA.__program_vars
-08 0x100001048 |---------#| 0x100001068 32 rw- 8.__DATA.__data
-=> 0x100000ef8 |----------| 0x100000ef7
EXPECT=''This one is due to platform environment variables differences: '%' vs '$' and making quotes a part of the variable:
[XX] db\cmd\cmd_system %ENV=33 !echo "$ENV" 1021
$ r2 -escr.utf8=0 -escr.color=0 -escr.interactive=0 -N -Q -i C:\Users\appveyor\AppData\Local\Temp\1\tmp-4216cx7o1ZK2rfXz.tmp -
%ENV=World
!echo $ENV
-World
+$ENV
EXPECT=<<EOF
$ENV
[XX] db\cmd\cmd_system !echo "test" 1007
$ r2 -escr.utf8=0 -escr.color=0 -escr.interactive=0 -N -Q -i C:\Users\appveyor\AppData\Local\Temp\1\tmp-4216aDuIzVAjSGDd.tmp -
!echo "test"
"!echo test"
+"test"
test
-test
EXPECT=<<EOF
"test"
testThis test looks very non-portable:
[XX] db\cmd\feat_redirect "pb>x;!wc -c x" 352
$ r2 -escr.utf8=0 -escr.color=0 -escr.interactive=0 -N -Q -i C:\Users\appveyor\AppData\Local\Temp\1\tmp-4216UsSgYZTIej4j.tmp malloc://1024
p8 10 > dump
!wc -c dump|sed 's/^[ \t]*//;s/[ \t]*$//'
p8 12800 > dump
!wc -c dump|sed 's/^[ \t]*//;s/[ \t]*$//'
!rm -f dump
-21
-25601
+21 dump
+25601 dumpThis discussion should be moved to r2r and issues open accordingly for every mentioned problems
XVilka
commented
5 years ago @a1ext note, that some !commands can be replaced with builtin alternatives:
!cat with cat!rm with rm!ls with `ls'!wc with ~?!wc -c with ~?.a1ext
commented
5 years ago This discussion should be moved to r2r and issues open accordingly for every mentioned problems
Why? Mostly problems fixes are in r2 side
a1ext
commented
5 years ago @a1ext note, that some
!commandscan be replaced with builtin alternatives:
!catwithcat!rmwithrm!lswith `ls'!wcwith~?!wc -cwith~?.
What about sed?
radare
commented
5 years ago Portability and performance
On 22 Mar 2019, at 09:09, a1ext notifications@github.com wrote:
This discussion should be moved to r2r and issues open accordingly for every mentioned problems
Why? Mostly problems fixes are in r2 side
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or mute the thread.
radare
commented
5 years ago Use ~ for sed but better not to have tests depending on it
On 22 Mar 2019, at 09:09, a1ext notifications@github.com wrote:
@a1ext note, that some !commands can be replaced with builtin alternatives:
!cat with cat !rm with rm !ls with `ls' !wc with ~? !wc -c with ~?. What about sed?
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or mute the thread.
Maijin
commented
5 years ago Because some are related to tests writing and not r2
a1ext
commented
5 years ago ok, could you transfer the issue to r2r?
Maijin
commented
5 years ago I cannot automatically because radare2 and radare2-regressions are not in radare2org #13263 . I will check later to do that manually.
XVilka
commented
5 years ago @a1ext is there any progress on this?
a1ext
commented
5 years ago @a1ext is there any progress on this?
Actually I have no time to work on any open-source at the moment, very busy. You can involve somebody else for the issue.
XVilka
commented
5 years ago @GustavoLCR you might want to take a look at those by the way.
radare
commented
5 years ago How many XX now?
XVilka
commented
5 years ago @radare 41, before the master was broken.
radare
commented
5 years ago status now?
GustavoLCR
commented
5 years ago 24 XX
GustavoLCR
commented
5 years ago Ok, I think that's it. Now waiting for @radare to release both r2r-js and r2pipe-js and see if everything is green.
radare
commented
5 years ago Both releases done
On 23 Jun 2019, at 03:31, GustavoLCR notifications@github.com wrote:
Ok, I think that's it. Now waiting for @radare to release both r2r-js and r2pipe-js and see if everything is green.
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or mute the thread.
XVilka
commented
5 years ago @GustavoLCR still 2 XX (see https://ci.appveyor.com/project/radare/radare2-shvdd/build/job/ea3a7x0f1tb9kj4x):
[XX] db\formats\mangling\mangling Abbreviation Of Repetead Names (Name with Template Arguments) demangler test 1510
$ r2 -escr.utf8=0 -escr.color=0 -escr.interactive=0 -N -Q -i C:\Users\appveyor\AppData\Local\Temp\1\tmp-4040b5VtciOCu83F -
"!rabin2 -D msvc ?xyz@?\$abc@V?\$def@H@@PAX@@YAXXZ"
-void __cdecl abc<class def<int>, void *>::xyz(void)
+?xyz@?\$abc@V?\$def@H@@PAX@@YAXXZ
EXPECT=<<EOF
?xyz@?\$abc@V?\$def@H@@PAX@@YAXXZ
[XX] db\formats\mangling\mangling Abbreviation Of Name with template demangler test 1466
$ r2 -escr.utf8=0 -escr.color=0 -escr.interactive=0 -N -Q -i C:\Users\appveyor\AppData\Local\Temp\1\tmp-4040HShhxRCI7n56 -
"!rabin2 -D msvc ??\$MyTemplateFunction@VClass1@@@Class1@@QAEXPAV0@@Z"
-public: void __thiscall Class1::MyTemplateFunction<class Class1>(class Class1 *)
+??\$MyTemplateFunction@VClass1@@@Class1@@QAEXPAV0@@Z
EXPECT=<<EOF
??\$MyTemplateFunction@VClass1@@@Class1@@QAEXPAV0@@ZAmazing job @GustavoLCR!
GustavoLCR
commented
5 years ago Amazing job @GustavoLCR!
Thanks!