Hi Andrew, hi lkml, On Tuesday 28 June 2005 13:00, you wrote: > From a maintainability POV it would be better to memset the whole array > beforehand - I changed the patch to do that) Yes, but then you should not assign 0xff to memset regions. > Signed-off-by: Zaur Kambarov <[email protected]> > Cc: <[email protected]? > Cc: Greg KH <[email protected]> > Signed-off-by: Andrew Morton <[email protected]> > --- > > drivers/usb/host/ohci-hub.c | 3 ++- > 1 files changed, 2 insertions(+), 1 deletion(-) > > diff -puN drivers/usb/host/ohci-hub.c~coverity-desc-bitmap-overrun-fix drivers/usb/host/ohci-hub.c > --- 25/drivers/usb/host/ohci-hub.c~coverity-desc-bitmap-overrun-fix 2005-06-24 22:11:00.000000000 -0700 > +++ 25-akpm/drivers/usb/host/ohci-hub.c 2005-06-24 22:19:48.000000000 -0700 > @@ -419,10 +419,11 @@ ohci_hub_descriptor ( > > /* two bitmaps: ports removable, and usb 1.0 legacy PortPwrCtrlMask */ > rh = roothub_b (ohci); > + memset(desc->bitmap, 0xff, sizeof(desc->bitmap)); > desc->bitmap [0] = rh & RH_B_DR; > if (ports > 7) { > desc->bitmap [1] = (rh & RH_B_DR) >> 8; > - desc->bitmap [2] = desc->bitmap [3] = 0xff; > + desc->bitmap [2] = 0xff; > } else > desc->bitmap [1] = 0xff; > } I would suggest: if (ports > 7) desc->bitmap[1] = (rh & RH_B_DR) >> 8 instead of the whole if construct. Regards Ingo Oeser
Attachment:
pgpk6vxIeWt6o.pgp
Description: PGP signature
- References:
- [no subject]
- From: "d binderman" <[email protected]>
- Re:
- From: Andrew Morton <[email protected]>
- [no subject]
- Prev by Date: Is Momentum Building for This Small-Cap?
- Next by Date: Re: kmalloc without GFP_xxx?
- Previous by thread: Re:
- Next by thread: Packet of death w/NFS+ULOGD+XFS under Kernel 2.6.x?
- Index(es):