Am 10.12.2014 um 01:21 schrieb Carlos E. R.:
On 2014-12-08 08:49, Bruno Friedmann wrote:
On Monday 08 December 2014 05.00:10 Carlos E. R. wrote:
How can dev fix a problem that can't be reproduced ... That's why adding people with same trouble or similar can really improve the situation
That's absurd. A good developer should be able to solve bugs just by looking at logs and code. Logs constitute "hard proof".
For instance, this one:
Date: Sat, 24 May 2014 23:40:06 +0000 Subject: [Bug 879778] New: Kernel crash on multiple file write on reiserfs GPT partition.
Took me many days to investigate and report, with some help from the mail list. No one has even commented or asked for more info or anything at all, after reporting in bugzilla...
And it is a BUG, because the kernel log says it is a bug:
[18892.976044] BUG: unable to handle kernel paging request at ffffc90012825250
There is no more confirmation needed. It IS a BUG.
Yes. But it's apparently a bad pointer dereference in the timer code => most likely a nasty race condition. That is usually not easily solved by "just looking at logs and code". Sorry. Now I know that reporting bugs against distro kernels is sometimes hard, simply because it is often the case that the distro's kernel maintainers are also busy with developing the mainline kernel. So the first thing I do is: I try if the bug is still present in the latest Kernel:Stable kernel, and if so, if it's still present in the latest Kernel:HEAD kernel. Often, if you can tell "this happens with 13.2 kernel, but no longer in the Kernel:Stable version", the maintainers will much more happily look into the differences and find something useful. Or, if it is still in Kernel:HEAD, then just go to the linux-kernel@vger.kernel.org list and complain about it there. Fortunately, the openSUSE kernel are close enough to mainline that quite often you can go to the upstream projects and just ask there, without them complaining that you absolutely need to reproduce this first with the hand-built Linus-tree-of-the-day. And yes, I am sometimes asking on opensuse-kernel instead of linux-kernel@vger, if I for example know that both people that are possible able to help me are actually kernel hackers employed by SUSE :-), see http://lists.opensuse.org/opensuse-kernel/2014-11/msg00059.html (the thread continues into December). I was pretty successful with getting my bugs fixed like that. Yes, it is a lot of work, but that's what we (as non-hardcore kernel hackers) just have to contribute to kernel development. If I don't want that, then I have to buy Windows (or SLES) and pay for support. For this particular bug, it might just be bad luck that reiserfs is no longer really actively maintained by anoyone? I don't know. Best regards, Stefan -- Stefan Seyfried "For a successful technology, reality must take precedence over public relations, for nature cannot be fooled." -- Richard Feynman -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org