Closed GoogleCodeExporter closed 8 years ago
just saying that it crashes won't help me figure out why. do you have any
information at all about where it crashes?
say, a stack trace at least. Do you know why the developers of
snappy-driver-installer believe this is a bug in libtorrent?
Original comment by arvid.no...@gmail.com
on 18 Mar 2015 at 1:38
do you have a .torrent file that reproduces the crash?
Original comment by arvid.no...@gmail.com
on 18 Mar 2015 at 1:38
client_test.exe is your example program from libtorrent-rasterbar-1.0.3.tar.
The torrent can be found at http://driveroff.net/SDI_Update.torrent but I think
it's irrelevant to the issue.
Thousands of people are using Snappy Driver Installer and no one reported about
this issue before. It must be something specific to this particular PC. I guess
we have to work with the end user and try to gather enough information to find
out why it happens but I don't know which facilities libtorrent has for
debugging crashes.
Snappy Driver Installer creates a backtrace.txt file when crashes but in this
case nothing is created. The latest messages from libtorrent are:
Torrent: dht_reply_alert | SDI_Update () received DHT peers: 61
Torrent: dht_reply_alert | SDI_Update () received DHT peers: 81
Torrent: dht_
[cut off]
Snappy Driver Installer didn't have a chance to flush the buffer, so the
logging ends abruptly.
Original comment by BadPointer0
on 18 Mar 2015 at 6:18
Sorry guys, I tried again this morning after rebooting and no worries now,
everything works fine.
However, I have not invented yesterday and I had the same problem on 2
computers, my Win7 x64 and a very new laptop with Win8.1 x64.
As it works this morning, I tried on a 3rd PC, also successfully.
It remained a mystery to me (same PCs, same config, no network concern) !!!,
but well, it works now.
Sorry again, keep us your good work :)
Original comment by Mateo.RO...@gmail.com
on 18 Mar 2015 at 8:46
Something must have been wrong with local network or ISP. Now that the issue
resolved itself, we'll have to wait till it resurfaces(if it ever happens
again).
It doesn't appear to be a racing condition issue because it was too consistent
for that.
Original comment by BadPointer0
on 18 Mar 2015 at 8:51
Just to let you know, as I had this concern, I downloaded, in the same time, as
other tests, the full download app with driverpacks (8.55 GB)
(http://snappy-driver-installer.sourceforge.net/SDI_Update.torrent) with
µtorrent with nothing special on that side.
I do not know if it is good it reappear again (not for me) to understand why. I
do not know more how to debug to find out what's wrong.
Thanks for your support :)
Original comment by Mateo.RO...@gmail.com
on 18 Mar 2015 at 9:35
[deleted comment]
[deleted comment]
I found SDI_R169D dump file on my win7 PC if it can help ?
https://dl.dropboxusercontent.com/u/16173277/SDI_R169D.exe.7924.7z
I do not have dump on the other Win8.1 PC, where I also tested libtorrent's
client_test alone.
Original comment by Mateo.RO...@gmail.com
on 18 Mar 2015 at 10:23
The app was complied by GCC, not Microsoft Visual Studio. I don't think this
dump file would be of any use.
It's not as easy to get backtrack on Windows as on Linux. It would be useful to
know which steps can be taken to debug libtorrent.
Original comment by BadPointer0
on 18 Mar 2015 at 10:25
for a second I thought that log was a stack trace, but it isn't. I still
haven't seen anything to suggest the bug is even in libtorrent.
PadPointer0, I would expect you to be able to debug your program, presumably
you built all the source. Probably this includes encoding version numbers in
dump files and storing source and .pdb file or equivalent, then load dump files
up in a debugger. I get the impression google breakpad is good at doing this.
Alternatively you could build with some amount of debug information enabled and
produce stack traces from a crash handler/signal handler and log it when
crashing.
Now, that doesn't mean it wasn't caused by a bug in libtorrent. For instance,
it's very possible that ones location in the DHT affects what kinds of packets
you're exposed to. Or that a specific peer in a swarm behaves strangely.
Neither of those cases would be easy for someone else to reproduce. If any of
you figure out where the crash happens, I'd be happy to take a look at at.
Original comment by arvid.no...@gmail.com
on 20 Mar 2015 at 1:12
Original issue reported on code.google.com by
Mateo.RO...@gmail.com
on 17 Mar 2015 at 1:35