On 9/27/05, Bernhard C. Schrenk <[email protected]> wrote: > I tested your previous version of this patch on an AMD64. I found one 64 bit > bug, crashing the kernel. After fixing it, it worked quite well. > > Please change in the file libata-core.c the line > > ap->hotplug_timer.data = (unsigned int)ap; > > to > > ap->hotplug_timer.data = (unsigned long)ap; Hey Bernhard, Thomas, Jeff, This bug was the product of my flagrant attempts to defy the C standard of sizeof(long) == sizeof(void *). Apparently, I have failed to slip it into the kernel once again. Super props go to me since I even have an Athlon 64 machine and didn't commit it to testing this. Jeff, please accept this revised patch in place of the original patch 2, the only difference being the change suggested above. Thanks to Thomas Lustig and Bernhard Schrenk for both pointing this out! Luke Kosewski
Attachment:
02-libata_hotswap_infrastructure-2.6.14-rc2-git5-2.diff
Description: Binary data
- References:
- [PATCH 2/3] Add disk hotswap support to libata RESEND #5
- From: Lukasz Kosewski <[email protected]>
- Re: [PATCH 2/3] Add disk hotswap support to libata RESEND #5
- From: "Bernhard C. Schrenk" <[email protected]>
- [PATCH 2/3] Add disk hotswap support to libata RESEND #5
- Prev by Date: Re: Strange disk corruption with Linux >= 2.6.13
- Next by Date: Re: [PATCH 2.6.14-rc2] fix incorrect mm->hiwater_vm and mm->hiwater_rss
- Previous by thread: Re: [PATCH 2/3] Add disk hotswap support to libata RESEND #5
- Next by thread: Re: [PATCH 2/3] Add disk hotswap support to libata RESEND #5
- Index(es):