Re: [Sdio-linux-devel] bug in sdio_memory_os.c?

[Paul L. <pl...@at...>]

I'm not sure if you didn't add in another truncation with the=20
(pInstance->Size* 1024).

Paul L

________________________________

From: Kari Nalli [mailto:ka...@gm...]=20
Sent: Wednesday, July 18, 2007 1:41 AM
To: sdi...@li...
Subject: [Sdio-linux-devel] bug in sdio_memory_os.c?



Hi,

I had some problems with 128MB SD memory card=20
----------------clip from log--------------
[  443.115173] +SDIO Memory Function: CreateDisk=20
[  443.119995] SDIO Memory Function: block size 512=20
[  443.124847 ] SDIO Memory Function: setting block size 512=20
[  443.130615] SDIO Memory Function: setting up DMA=20
[  443.135681] SDIO Memory Function: CreateDisk: size 245760 (Size=20
123008, FileSysBlockSize 512)=20

=2E.. and later...=20

[  445.515045] -SDIO Memory Function: AsyncCompletionLast=20
[  445.521118]  p1=20
[  445.522979]  sdmem0: p1 exceeds device capacity
 ----------------clip from log--------------

Now looking at code on sdio_memory_os.c:=20
set_capacity(pInstance->Config.pGenDisk,pInstance->Size/pInstance-=20
>FileSysBlockSize*1024);
This causes rounding errors and disk size 245760

after changing code to
set_capacity(pInstance-> Config.pGenDisk , (pInstance->Size*=20
1024)/pInstance->FileSysBlockSize);

The 128 MB card is functional (card has only one partition) and=20
/sys/block/sdmem0/sdmem0p1# more size -> 245919 sectors which is bigger=20
than originally calculated (245760)=20

So is this bug on driver or is there some reason way calculation was=20
originally done way it was?

BR, Kari