fche changed the topic of #systemtap to: http://sourceware.org/systemtap; email systemtap@sourceware.org if answers here not timely, conversations may be logged
hpt has joined #systemtap
sfink has quit [Ping timeout: 255 seconds]
hpt has quit [Quit: Lost terminal]
hpt has joined #systemtap
sfink has joined #systemtap
sfink has quit [Read error: Connection reset by peer]
sfink has joined #systemtap
sfink has quit [Ping timeout: 272 seconds]
hpt has quit [Ping timeout: 240 seconds]
sfink has joined #systemtap
hpt has joined #systemtap
hpt has quit [Quit: leaving]
hpt has joined #systemtap
hpt has quit [Ping timeout: 250 seconds]
hpt has joined #systemtap
nkambo1 has quit [Ping timeout: 256 seconds]
ego has joined #systemtap
ananth has joined #systemtap
naveen has joined #systemtap
nkambo1 has joined #systemtap
naveen1 has joined #systemtap
naveen has quit [Ping timeout: 265 seconds]
ego has quit [Ping timeout: 240 seconds]
Humble has joined #systemtap
irker745 has quit [Quit: transmission timeout]
nkambo1 has quit [Ping timeout: 260 seconds]
hkshaw has joined #systemtap
ego has joined #systemtap
hchiramm has joined #systemtap
ego has quit [Client Quit]
ego has joined #systemtap
naveen1 has quit [Quit: Leaving.]
naveen has joined #systemtap
nkambo1 has joined #systemtap
srikar_away is now known as srikar
naveen has quit [Quit: Leaving.]
naveen has joined #systemtap
naveen has quit [Quit: Leaving.]
naveen has joined #systemtap
ego has quit [Ping timeout: 240 seconds]
ego has joined #systemtap
ego has quit [Ping timeout: 255 seconds]
ego has joined #systemtap
nkambo1 has quit [Ping timeout: 272 seconds]
hpt has quit [Ping timeout: 272 seconds]
nkambo1 has joined #systemtap
nkambo1 has quit [Ping timeout: 272 seconds]
naveen has quit [Ping timeout: 264 seconds]
nkambo1 has joined #systemtap
naveen has joined #systemtap
ego has quit [Ping timeout: 256 seconds]
naveen has quit [Ping timeout: 256 seconds]
wakatana2 has joined #systemtap
wakatana2 has quit [Max SendQ exceeded]
naveen has joined #systemtap
ego has joined #systemtap
darvon has quit [Ping timeout: 250 seconds]
modem has quit [Ping timeout: 240 seconds]
darvon has joined #systemtap
modem has joined #systemtap
naveen has quit [Quit: Leaving.]
ananth has quit [Quit: Leaving]
ego has quit [Quit: Leaving]
hkshaw has quit [Ping timeout: 246 seconds]
Empyrium has quit [Remote host closed the connection]
mjw has joined #systemtap
hkshaw has joined #systemtap
flip214 has joined #systemtap
<flip214>
Hi.
<flip214>
some if() in a kernel modules is in assembler "move %rsp(0x20), edx; test edx,edx" etc.
<flip214>
sadly the variable isn't seen by systemtap at this point; so I'd either need to access "%rsp(0x20)" or edx a few bytes later.
<flip214>
I haven't found out how to says function("xxx+offset"), though. is that not supported yet?
<flip214>
as it's a module, using the absolute address isn't that nice, too...
<flip214>
alternatively, how can I fetch the value from the stack?
<flip214>
I'm using systemtap=2.9-2, btw.
hkshaw has quit [Ping timeout: 240 seconds]
hkshaw has joined #systemtap
srikar is now known as srikar_away
nkambo1 has quit [Ping timeout: 255 seconds]
guy_ has left #systemtap [#systemtap]
wcohen has quit [Remote host closed the connection]
srikar_away is now known as srikar
fche2 is now known as fche
<fche>
flip214, you can use .statement probes for the interior of a function
mbenitez has joined #systemtap
mbenitez has quit [Changing host]
mbenitez has joined #systemtap
<fche>
for values from the stack, something like kernel_long(register("sp")) kind of thing
<flip214>
fche: yes. but if it's a module, I have to manually add the function address (from /proc/ksyms) and the offset - and that might change every time I load the module...
<flip214>
I'll try kernel_long
<fche>
no debuginfo for the module?
<flip214>
oh yes, but the instruction is the second one for this source line.
<flip214>
so I need to use .statement().absolute
<flip214>
and that doesn't do arithmetic, does it?
<fche>
nope.
<fche>
is the next statement/source-line too late?
<flip214>
yeah.
<flip214>
the source line points to the "mov %rsp(0x20), edx" but I'd like the register value after the load
<fche>
you could probe the source line itself and emulate the stack load
<fche>
(or even dump the stack with something like printf("%*M", number_of_bytes, register("sp"))
<flip214>
yeah, I load from the stack now, via kernel_load that you told me about.
drsmith has joined #systemtap
srikar is now known as srikar_away
tromey has joined #systemtap
srikar_away is now known as srikar
wcohen has joined #systemtap
<mjw>
fche, hmmm, those Debian8 warnings are kind of bad. I don't have a Debian8 install atm. Still on 7. I wonder if this is also the silly compressed debuginfo thing again :{
irker128 has joined #systemtap
<irker128>
systemtap: zhouwj-fnst systemtap.git:refs/heads/master * release-2.9-206-g86cfd04 / testsuite/config/unix.exp testsuite/lib/gen_load.exp testsuite/systemtap.stress/all_kernel_functions.exp testsuite/systemtap.stress/current.exp: Move testsuite 'genload' function to its own file. http://tinyurl.com/zpcz9jx
<fche>
mjw, I gather it mostly works
<mjw>
I wanted to say that updating to elfutils 0.165 would at least work around that. But that isn't true for the unwinder because that does most things by hand.
<mjw>
If it is though it shouldn't be too hard to fix using elfutils 0.165 because then you just have to be sure to call elf_compress first on the section
<mjw>
hmmm, this is for eh_frame. Which makes almost everything I said above wrong. .eh_frame is an allocated section and so cannot be compressed
<mjw>
odd
<fche>
one of the warnings was for unwiding-through-self
<mjw>
I'll have to install Debian 8 one of these days anyway. Will try to play with it then.
<mjw>
unwinding-through-self is so cool!
brolley has joined #systemtap
Empyrium has joined #systemtap
hchiramm has quit [Ping timeout: 240 seconds]
Humble has quit [Ping timeout: 264 seconds]
srikar is now known as srikar_away
srikar_away is now known as srikar
nkambo has joined #systemtap
hkshaw has quit [Quit: Leaving.]
<irker128>
systemtap: dsmith systemtap.git:refs/heads/master * release-2.9-207-g82343d7 / tapsets.cxx: Update tracepoint support for 4.4.0 aarch64 kernels. http://tinyurl.com/gv46rwh
nkambo has quit [Ping timeout: 272 seconds]
mbenitez has quit [Quit: To office]
nkambo has joined #systemtap
mbenitez has joined #systemtap
mbenitez has joined #systemtap
wakatana2 has joined #systemtap
cbzx has joined #systemtap
wakatana2 has quit [Max SendQ exceeded]
srikar is now known as srikar_away
drsmith has left #systemtap [#systemtap]
nkambo1 has joined #systemtap
nkambo has quit [Ping timeout: 250 seconds]
cbzx has quit [Quit: Konversation terminated!]
tromey has quit [Quit: ERC (IRC client for Emacs 25.0.50.2)]