pmladek
Prague
Grub2: make it easy to select the default boot target
an idea by pmladek
"grub2-once" allows to print a numbered list of available targets using --list option. It allows to select the boot target using the number. It would be great to add this functionality to the "grub2-set-default" command as well.
Disk operations should not stale the system
an idea by pmladek
It happens from time to time that my system is less responsible. It is usually caused by extensive disk operations, e.g. searching disk for something, moving a lot of data. A solution would be to run these problematic tasks with ionice. But I think that desktop user should not need to take care of this. Also some of the operations are even caused by system tools and normal user has problems to affect it. I wonder if we could somehow improve the desktop setting or it it would even need some hacks into the scheduler. I have just heard that this should be fined in the last kernels. But I also heard that it still does not work well, for example when copying from/to USB memory stick.
kGraft: allow to handle more patches in parallel
a project by pmladek
Current kGraft implementation allows to apply new patch only when the previous one is applied to the whole system. Every task is handled separately and it needs to wait for a safe place to switch. It might take a while, especially when the task is sleeping. Therefore it would be great to allow to apply more patches in parallel. It is not trivial because different tasks might be in different world, so the slow stub has to be clever and choose the right one without taking any lock...
kGraft: allow to define safe place where the code can be switched in kthreads
a project by pmladek
<p>kGraft allows to patch kernel at runtime. It implements a consistency model that allows to modify semantic of functions. e.g. fix lock order. For this, we need to find a safe place when a process/thread might start using the new code.</p> <p>Userspace tasks are switched on the kernel boundary when syscall is called/returns or when a signal is proceed.</p>
Simplify implementations of lockless printk ring buffer
an invention by pmladek
There are attempts to solve some kernel deadlock with using lockless printk ringbuffer. The proposed implementation is pretty complex (6 stages, 6 write and 6 read barriers, two buffers, entries linked via list, ...) I have a idea how to make it easier with tracking the state and sequence number in one atomic variable. It might allow to remove the lists and many barriers. It is possible that it will just not work. Let's see.
Add support for the new lockless printk ringbuffer into crash-python
an idea by pmladek
printk() is function that is used to print messages in the kernel code. The messages are stored into an internal buffer and show on the registered consoles. They could be read and stored by userspace when the system is running. The messages might not reach the console or userpace when the system crashes. One way to see them is reading them from a kernel crashump. crash-python is a semantic debugger for the Linux kernel. It is a very powerful tool for analyzing kernel crashdumps.
Improve debugging of kernel workques
a project by pmladek
Project Description
Linus kernel has a subsystem called workqueues. It provides pool of generic
Looking for projects around:
Nothing at the moment
Activity