Development mailing list

Syndicate content
Archive of posts for haiku-development at FreeLists
Updated: 29 min 42 sec ago

[haiku-development] Re: killing atomic syscalls? (Ithamar Adema)

Mon, 2015-03-16 12:45
Hi, On Thu, Feb 26, 2015 at 10:30 AM, François Revol revol@xxxxxxx wrote: On ARM currently some variants still need syscalls for the 64bit atomics. I'm not sure which though, likely the asm implementations are inside some #ifdef ? ARM11 need both 32 and 64 bit versions to be sys calls, ARM11 needs the ...
Categories: Development

[haiku-development] Re: killing atomic syscalls? (François Revol)

Mon, 2015-03-16 12:45
On 26/02/2015 08:27, Axel Dörfler wrote: Am 26/02/2015 um 05:00 schrieb kallisti5: It looks like after Paweł's C++11 work, we don't need the atomic syscalls anymore? There are architectures that need syscalls to implement the atomic functions properly. Those are the only ones that should use them. On ARM currently some variants still need syscalls for the 64bit atomics. I'm not sure which though, likely ...
Categories: Development

[haiku-development] Re: killing atomic syscalls? (Axel Dörfler)

Mon, 2015-03-16 12:45
Am 26/02/2015 um 05:00 schrieb kallisti5: It looks like after Paweł's C++11 work, we don't need the atomic syscalls anymore? There are architectures that need syscalls to implement the atomic functions properly. Those are the only ones that should use them. ...
Categories: Development

[haiku-development] Re: Haiku Userland on Non-Haiku Kernel (Stephan)

Mon, 2015-03-16 12:45
Hi! For those who are still interested: I have been very very busy during the last days so there is not much progress. Appserver and registrar are running on NetBSD console and (sometimes) X. The appserver is not quite stable. When I launch an application, registrar notices that but crashes immediately afterwards. I need a full debug build of NetBSD (especially libc) to get that properly debugged. ...
Categories: Development

[haiku-development] Re: killing atomic syscalls? (Michael Crawford)

Mon, 2015-03-16 12:45
I'm not familiar with the code in question, but in general if something can be done in userspace rather than in a system call, it would be a huge win. System calls are in general implemented by executing an instruction which is illegal in userspace. An exception is taken, _all_ of the registers are saved on the stack, a dispatch table is used to determine what kernel function to call to carry out that system call's work. Upon return, all those registers are restored from the exception stack. ...
Categories: Development

[haiku-development] killing atomic syscalls? (kallisti5)

Mon, 2015-03-16 12:45
It looks like after Paweł's C++11 work, we don't need the atomic syscalls anymore? The main reason for asking is I attempted an arm (Raspberry Pi + u-boot) bootstrap and ran into: . ...
Categories: Development

[haiku-development] Re: Questions about code reviews and static analysis tools for TU Delft research (Radjino Bholanath)

Sat, 2015-03-14 14:45
Thanks for answering Augustin, I appreciate it. - Radjino ________________________________________ From: haiku-development-bounce@xxxxxxxxxxxxx [haiku-development-bounce@xxxxxxxxxxxxx] on behalf of Augustin Cavalier [waddlesplash@xxxxxxxxx] Sent: Tuesday, February 24, 2015 3:52 PM To: haiku-development@xxxxxxxxxxxxx Subject: [haiku-development] Re: Questions about code reviews and static ...
Categories: Development

[haiku-development] Re: Questions about code reviews and static analysis tools for TU Delft research (Augustin Cavalier)

Wed, 2015-03-11 07:45
On 2/24/2015 5:53 AM, Radjino Bholanath wrote: 1. Do all developers (contributors and core developers) have to submit a code review for every change? I’m asking because many projects only review changes made by contributors. No, we don't. It has been discussed in the past but the consensus was that the developers have too little time to make reviews mandatory before pushes. ...
Categories: Development

[haiku-development] Questions about code reviews and static analysis tools for TU Delft research (Radjino Bholanath)

Mon, 2015-03-09 21:45
Hi everyone, I'm doing research on code reviews and static analysis tools at the SERG group () of the Delft University of Technology. Currently, we want to give an overview of the usage of code review and static analysis tools in open source projects. Therefore, I would be very happy to know a little bit more about how code reviews are used in Haiku and if (and maybe how) static analysis tools are used. I have a couple of questions for anyone willing to answer: ...
Categories: Development

[haiku-development] Re: Review needed for Sector Size not being reported correctly bug (Ingo Weinhold)

Mon, 2015-03-09 15:45
On 02/22/2015 08:51 PM, Kushal Singh wrote: On Sun Feb 22 2015 at 8:55:14 PM Ingo Weinhold ingo_weinhold@xxxxxx wrote: Long story short, the file systems setting partition_data::block_size to their content block size is correct, or IOW your patch 0001 is incorrect I have only glanced over the comments in the ticket. From that and the ...
Categories: Development

[haiku-development] Re: Review needed for Sector Size not being reported correctly bug (Kushal Singh)

Sat, 2015-03-07 21:45
Also one more point that I missed in my most recent mail On Sun Feb 22 2015 at 8:55:14 PM Ingo Weinhold ingo_weinhold@xxxxxx wrote: The block_size field doesn't have a corresponding content_block_size field. The rationale for this is that partitioning systems don't specify block sizes for partitions they define. They just have a block size they use for defining/interpreting partition offsets and sizes. So, in fact, they have a content block size, just as file systems have a content block size. Since there is no other block size for a partition, the ...
Categories: Development

[haiku-development] Re: Review needed for Sector Size not being reported correctly bug (Kushal Singh)

Thu, 2015-03-05 14:45
On Sun Feb 22 2015 at 8:55:14 PM Ingo Weinhold ingo_weinhold@xxxxxx wrote: Long story short, the file systems setting partition_data::block_size to their content block size is correct, or IOW your patch 0001 is incorrect I have only glanced over the comments in the ticket. From that and the not very helpful ticket description unfortunately I cannot say what the actual problem is that your patch 0001 is meant to fix. Your comment 17 [1] seems to imply that when uninitializing/re-initializing a partition the block_size field isn't reset correctly. ...
Categories: Development

[haiku-development] Re: Review needed for Sector Size not being reported correctly bug (Ingo Weinhold)

Thu, 2015-03-05 12:45
On 02/22/2015 02:03 PM, Kushal Singh wrote: I was working on . Initially it involved making sure that partitions with size 2TB was not allowed. However later on it was found out that somewhere in the code file system block size and block device block size (aka sector) were being confused , as the BlockSize returned file system block size instead of sector size. ...
Categories: Development

[haiku-development] Review needed for Sector Size not being reported correctly bug (Kushal Singh)

Thu, 2015-03-05 04:45
Hi, I was working on . Initially it involved making sure that partitions with size 2TB was not allowed. However later on it was found out that somewhere in the code file system block size and block device block size (aka sector) were being confused , as the BlockSize returned file system block size instead of sector size. I have created a patch for the issue . Here is the link . ...
Categories: Development

[haiku-development] Re: PM Mount Points (Ingo Weinhold)

Thu, 2015-03-05 04:45
On 02/20/2015 11:53 AM, Axel Dörfler wrote: Maybe we should investigate the amount of work this causes, and then vote what we should try to pursue (if nobody is found to carry out the change, it's obviously not that important anyway). The obvious changes should be: - Change the paths find_path() and friends return - Eventually adapt the boot loader to still find the loader ...
Categories: Development

[haiku-development] Re: PM Mount Points (Ingo Weinhold)

Wed, 2015-03-04 16:45
On 02/20/2015 11:32 AM, Stephan Aßmus wrote: Am 20.02.2015 um 11:16 schrieb Ingo Weinhold: The argument isn't so much that users may be confused -- as you write, depending on their level they won't be or they won't even know -- it's that this is a weird way of organizing things: Virtually all software lives in a subdirectory while the prominent places (/system/* and ~/config/*) are essentially empty. ...
Categories: Development

[haiku-development] Re: PM Mount Points (pete . goodeve)

Wed, 2015-03-04 00:45
On Fri, Feb 20, 2015 at 01:27:55PM -0800, Urias McCullough wrote: On Fri, Feb 20, 2015 at 1:23 PM, John Scipione jscipione@xxxxxxxxx wrote: my preference: /boot/system (non-packaged) /boot/system/packages (packaged) /boot/home/config (non-packaged) /boot/home/config/packages (packaged) that way /boot/system exactly matches /boot/home/config. packages go in ...
Categories: Development

[haiku-development] Re: PM Mount Points (John Scipione)

Tue, 2015-03-03 22:45
On Feb 20, 2015, at 4:27 PM, Urias McCullough umccullough@xxxxxxxxx wrote: On Fri, Feb 20, 2015 at 1:23 PM, John Scipione jscipione@xxxxxxxxx wrote: my preference: /boot/system (non-packaged) /boot/system/packages (packaged) /boot/home/config (non-packaged) /boot/home/config/packages (packaged) ...
Categories: Development

[haiku-development] Re: PM Mount Points (Urias McCullough)

Tue, 2015-03-03 16:45
On Fri, Feb 20, 2015 at 1:23 PM, John Scipione jscipione@xxxxxxxxx wrote: my preference: /boot/system (non-packaged) /boot/system/packages (packaged) /boot/home/config (non-packaged) /boot/home/config/packages (packaged) that way /boot/system exactly matches /boot/home/config. packages go in read-only packages directories, otherwise the layout resembles on pre-PM ...
Categories: Development

[haiku-development] Re: PM Mount Points (John Scipione)

Tue, 2015-03-03 14:45
On Feb 20, 2015, at 3:59 PM, pete.goodeve@xxxxxxxxxxxx wrote: On Fri, Feb 20, 2015 at 09:29:42AM +0100, Axel Dörfler wrote: Am 19/02/2015 um 21:29 schrieb Adrien Destugues: And what I'd prefer in that case (I'm fine with either this or the current one): /boot/system (not packages) /boot/system/packaged (packages) ~/config/packaged (packages) ~/config/ (not packages) ...
Categories: Development
<