Using a real harddisk

[Sebastian Sumpf]

Hi,

On 05/11/2014 09:17 PM, w_schmidt at ...181... wrote:
> Hi,
> 
> I have a few stupid questions regarding the use of real hardware.
> 
> I want to access files from a harddisk using a block cache. The partition 
> would be /sda7/test in ext2 format. (and for example file1.txt) I tried to 
> start nova directly form disc - works with mouse & keyboard now.
> 
> With the example file in /dde_rump/run/rump2_ext2.run I do not see how I 
> could use an existing harddrive.
> 
> I understand the script as following:
> First a bin/ext2.raw file is created and then a file system is created with 
> mke2fs.
> The ram_blk points to the file. How can it be pointed to a drive like /sda7 
> instead?
> Is it even possible now to use a real harddsic file system? Or would i need 
> to create a file like in the example (only permanent) and put the file 
> together with the rest of the program in grub so that nova boots with 
> providing the file?

Assuming your machine is an x86 derivative and supports AHCI, you could
use Genode's AHCI driver, which will expose a block sesssion (see:
os/src/drivers/ahci/README). In order to access the partitions on your
disk, a server called 'part_blk' is required (see:
os/src/server/part_blk/README for its configuration). So, you would have
to remove 'ram_blk' and add 'ahci_drv', 'part_blk' (routed to
'ahci_drv'), adjust 'rump_fs' to be routed to 'part_blk' partition 7.

> The second part of the question is according to things needed to execute the 
> rump_ext2.run script.
> As i tried to create prepare PKG=libs i needded to install:  subversion, 
> flex, bison
> Are they missing in the build tools or is no specific version needed?

There should be no specific version needed.

> The next questions are regarding the block cache. If I try to execute the 
> script, i get a Tests finished succesful, but a lot of messages like
> [init] Cannot respond to resource request - out of memory
> [init ->blk_cache] could not expand dataspace pool

The cache is greedy and tries to get as much memory from its parent as
possible. By the way, rump has a build in block cache also.

> To be honest I do not see in the configuration at all how it should work.
> If i understand it correctly a qemu with 64 MB RAM is started, and 
> test-blk-cli is given a quantum of 2GB?

That only means that test-blk-cli will get all the remaining RAM.

> The test-blk-cli gets a route for Block to the Cache and the Cache to 
> Blksrv. What I don't understand with the sizes is: Why has the client the 
> largest size and not the smallest? If it has the largest it could just read 
> complete files and keep it in memory?

I think the author just did not want to calculate the amount of RAM
required for the test program.

> I tried the following: I copied the rump_ext2.run script and created an 
> entry with start name blk_cache and put a route from rump_fs to Block child 
> blk_cache, from blk_Cache as child ram_blk.
> I started it and then see that the script said that blk_cache was not found, 
> but it run succesful.
> 
> My question is therefore: if a service name is in multiple clients and one 
> client in a route is not available, is automatically another client with 
> same service name used? Or have i forgot something?

It worked because 'blk_cache' wasn't started, and therefore only one
block session was announced ('ram_blk'). If the block cache had been
started and 'test-libc_vfs' didn't have a route entry for the block
service, init would complain about ambiguous routes to service "Block".

> Afterwards i included in the set build_components section server/blk_cache 
> (after drivers/timer) and entered as boot module blk_cache.  Is this the 
> correct way of including the cache?

Yes it is!

> The result of this was that I got an output with: [init -> blk_cache] 
> updating quota for SIGNAL session, an init rump_fs upgrade quota output and 
> afterwards and a long stop after init->rump_fs Backend::Backend() Backend 
> blk_size 512 Afterwards the script stopped with   Test execution timeout. 
> with error 254) What could I have done wrong with the configuration of this 
> scenario?

Ok, I will try to reproduce the behavior.

Regards,

Sebastian


-- 
Sebastian Sumpf
Genode Labs

http://www.genode-labs.com · http://genode.org

Genode Labs GmbH · Amtsgericht Dresden · HRB 28424 · Sitz Dresden
Geschäftsführer: Dr.-Ing. Norman Feske, Christian Helmuth