linux-kernel - Re: [PATCH] PCI: brcmstb: Fix control flow issue

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <CA+-6iNw4TiKtH4OWiAHpGC1KK_R3bkBQoBqxdNF2oh1vDrh8fA@mail.gmail.com>
Date: Wed, 11 Sep 2024 11:02:28 -0400
From: Jim Quinlan <james.quinlan@...adcom.com>
To: Qianqiang Liu <qianqiang.liu@....com>
Cc: Bjorn Helgaas <helgaas@...nel.org>, bcm-kernel-feedback-list@...adcom.com, 
	florian.fainelli@...adcom.com, linux-pci@...r.kernel.org, 
	linux-kernel@...r.kernel.org
Subject: Re: [PATCH] PCI: brcmstb: Fix control flow issue

On Wed, Sep 11, 2024 at 9:56 AM Qianqiang Liu <qianqiang.liu@....com> wrote:
>
> On Wed, Sep 11, 2024 at 08:44:36AM -0500, Bjorn Helgaas wrote:
> > On Wed, Sep 11, 2024 at 10:50:59AM +0800, Qianqiang Liu wrote:
> > > The type of "num_inbound_wins" is "u8", so the less-than-zero
> > > comparison of an unsigned value is never true.
> >
> > I think this was fixed slightly differently but with the same effect;
> > please check this to make sure:
> >
> > https://git.kernel.org/pub/scm/linux/kernel/git/pci/pci.git/tree/drivers/pci/controller/pcie-brcmstb.c?h=controller/brcmstb#n1034
> >
> > > ---
> > >  drivers/pci/controller/pcie-brcmstb.c | 8 +++++---
> > >  1 file changed, 5 insertions(+), 3 deletions(-)
> > >
> > > diff --git a/drivers/pci/controller/pcie-brcmstb.c b/drivers/pci/controller/pcie-brcmstb.c
> > > index 55311dc47615..3e4572c3eeb1 100644
> > > --- a/drivers/pci/controller/pcie-brcmstb.c
> > > +++ b/drivers/pci/controller/pcie-brcmstb.c
> > > @@ -1090,9 +1090,11 @@ static int brcm_pcie_setup(struct brcm_pcie *pcie)
> > >     u32p_replace_bits(&tmp, 1, PCIE_MISC_MISC_CTRL_PCIE_RCB_64B_MODE_MASK);
> > >     writel(tmp, base + PCIE_MISC_MISC_CTRL);
> > >
> > > -   num_inbound_wins = brcm_pcie_get_inbound_wins(pcie, inbound_wins);
> > > -   if (num_inbound_wins < 0)
> > > -           return num_inbound_wins;
> > > +   ret = brcm_pcie_get_inbound_wins(pcie, inbound_wins);
> > > +   if (ret < 0)
> > > +           return ret;
> > > +
> > > +   num_inbound_wins = (u8)ret;
> > >
> > >     set_inbound_win_registers(pcie, inbound_wins, num_inbound_wins);
> > >
> > > --
> > > 2.39.2
> > >
>
> Yes, they have the same effect.

Qianqiang's fixup has the slight advantage of keeping the var a u8,
which is what a reviewer wanted in the first place.
I found it surprising that  "if (u8var < 0)" does not cause a compiler
warning, but it appears that W=1 does not include "-Wtype-limits".
That's no excuse for this mistake, of course :-(


Thanks and regards,
Jim Quinlan Broadcom STB/CM

>
> --
> Best,
> Qianqiang Liu
>

Download attachment "smime.p7s" of type "application/pkcs7-signature" (4210 bytes)