[haiku-bugs] Re: [Haiku] #8990: intel partiton addon allows creating partitions > 2TB (easy)
- From: "pdziepak" <trac@xxxxxxxxxxxx>
- Date: Mon, 26 Jan 2015 11:49:58 -0000
#8990: intel partiton addon allows creating partitions > 2TB (easy)
------------------------------------------+----------------------------
Reporter: luroh | Owner: bonefish
Type: bug | Status: new
Priority: normal | Milestone: R1
Component: Partitioning Systems/Intel | Version: R1/Development
Resolution: | Keywords:
Blocked By: | Blocking:
Has a Patch: 1 | Platform: All
------------------------------------------+----------------------------
Comment (by pdziepak):
Replying to [comment:13 kushalsingh007]:
> Replying to [comment:12 pdziepak]:
> >
> > > 1.Using {{{partition->Size() / partition->BlockSize() <
UINT32_MAX}}} causes the condition to fail.
> > > For instance, If BlockSize = 2048 ={{{2^11}}}.Then
partition->Size() < {{{2^32}}}*{{{2^11}}} or {{{2^43}}} which is 8TB.
> > > Thus clearly the condition fails. Also there can be overflow error
using the previous condition. Moreover the offset wasn't been checked
anywhere.
> >
> > Well, I am getting a bit confused here. When reading the MBR add-on
code I got the impression that `BlockSize()` is supposed to return the
sector size (apparently the original author of the check in question
assumed the same). In such case there is nothing wrong in the scenario you
presented. 2TB partition size limit applies only when sector size 512
because that limit originates from the 32 bit size of "partition size in
sectors" value. If the sector size is 2kB nothing prevents MBR from
supporting 8TB partitions. Have you actually seen `BlockSize()` returning
something else than 512 on a disk with 512 byte sectors?
>
> Yes , while testing I found the blocksize to be 2048 when sector size
was 512. It depends on the value that has been initialized by you.
What is this "block size" then? The only "blocks" at this abstraction
level are sectors.
> >
> > >
> > > 2.Yeah I assumed the sector size to be 512 , the max limit is
{{{UINT32_MAX * SECTOR_SIZE}}}, but then since MBR assumes 512 bytes
sectors , so I think that's fine.
> >
> > If I understand the MBR code correctly `BlockSize()` returns size of
the sector. So the only change required is checking whether the offset is
`< BlockSize() * UINT32_MAX`.
> >
> But then again BlockSize() returns the size of the block (512/2048/4096
etc) not the size of the sector. Had it returned the size of the sector ,
your assumption would have been true.
Commit 5f1b31debdf97da331be6e97a6c2bc5899e2423a suggests that my
assumption is indeed true and `BlockSize()` should always return the size
of a sector. If it doesn't then we probably have another bug.
--
Ticket URL: <https://dev.haiku-os.org/ticket/8990#comment:15>
Haiku <https://dev.haiku-os.org>
Haiku - the operating system.
Other related posts:
