linux-kernel - Re: [PATCH] PCI: Mediatek: Use resource

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]

Date:   Thu, 31 Jan 2019 15:52:36 +0800
From:   Honghui Zhang <honghui.zhang@...iatek.com>
To:     Bjorn Helgaas <helgaas@...nel.org>
CC:     <lorenzo.pieralisi@....com>,
        <linux-arm-kernel@...ts.infradead.org>,
        <linux-mediatek@...ts.infradead.org>, <linux-pci@...r.kernel.org>,
        <linux-kernel@...r.kernel.org>, <ryder.lee@...iatek.com>,
        <youlin.pei@...iatek.com>, <jianjun.wang@...iatek.com>
Subject: Re: [PATCH] PCI: Mediatek: Use resource_size function on resource
 object

On Thu, 2019-01-31 at 11:03 +0800, Honghui Zhang wrote:
> On Wed, 2019-01-30 at 09:49 -0600, Bjorn Helgaas wrote:
> > On Wed, Jan 02, 2019 at 02:03:53PM +0800, honghui.zhang@...iatek.com wrote:
> > > From: Honghui Zhang <honghui.zhang@...iatek.com>
> > > 
> > > drivers/pci/pcie-mediatek.c:720:13-16: WARNING: Suspicious code. resource_size is maybe missing with mem
> > > 
> > > Generated by: scripts/coccinelle/api/resource_size.cocci
> > > 
> > > Signed-off-by: Honghui Zhang <honghui.zhang@...iatek.com>
> > > ---
> > >  drivers/pci/controller/pcie-mediatek.c | 4 +---
> > >  1 file changed, 1 insertion(+), 3 deletions(-)
> > > 
> > > diff --git a/drivers/pci/controller/pcie-mediatek.c b/drivers/pci/controller/pcie-mediatek.c
> > > index e307166..0168376 100644
> > > --- a/drivers/pci/controller/pcie-mediatek.c
> > > +++ b/drivers/pci/controller/pcie-mediatek.c
> > > @@ -654,7 +654,6 @@ static int mtk_pcie_startup_port_v2(struct mtk_pcie_port *port)
> > >  	struct resource *mem = &pcie->mem;
> > >  	const struct mtk_pcie_soc *soc = port->pcie->soc;
> > >  	u32 val;
> > > -	size_t size;
> > >  	int err;
> > >  
> > >  	/* MT7622 platforms need to enable LTSSM and ASPM from PCIe subsys */
> > > @@ -706,8 +705,7 @@ static int mtk_pcie_startup_port_v2(struct mtk_pcie_port *port)
> > >  		mtk_pcie_enable_msi(port);
> > >  
> > >  	/* Set AHB to PCIe translation windows */
> > > -	size = mem->end - mem->start;
> > > -	val = lower_32_bits(mem->start) | AHB2PCIE_SIZE(fls(size));
> > > +	val = lower_32_bits(mem->start) | AHB2PCIE_SIZE(fls(resource_size(mem)));
> > 
> > This is actually a fairly interesting change because it effectively
> > changes this:
> > 
> >   fls(mem->end - mem->start)
> > 
> > to this:
> > 
> >   fls(mem->end - mem->start + 1)
> > 
> > And mem->end is the last valid address, so it changes something like
> > this:
> > 
> >   fls(0xffff)         # == 15
> > 
> > to this:
> > 
> >   fls(0x10000)        # == 16
> > 
> > So while this *looks* like a trivial warning fix, it likely fixes an
> > important bug, and it's worth pointing out what that bug is in the
> > changelog.
> > 
> > >  	writel(val, port->base + PCIE_AHB_TRANS_BASE0_L);
> > >  
> > >  	val = upper_32_bits(mem->start);
> 
> This size will set the MMIO size, which means that the RC will translate
> the MMIO access from mem->start to mem->end.
> The real MMIO size is specified in devicetree, which is 0x1000_0000 for
> both mt2712 and mt7622.
> 
> This change make the size from fls(0x1000_0000) to fls(0x1000_0001), not
> really change the values.
> 
> I will update the commit message and add the information mentioned
> above.
> 
> Thanks for your kindly review.

I was wrong, after take a look at the resource parser function, that it
will initialize the res->end as res->start + res->size - 1.

But this change is still Ok since it will change the MMIO from
fls(0xfff_ffff) to fls(0x1000_0000), this just enlarge the MMIO
translate window size. The HW assigned MMIO is 0x1000_0000, but original
code set this translate window to fls(0xfff_ffff) = 0x800_0000 is fine
in most case since all the EP device we connect never asked a MMIO
window bigger than 0x800_0000. (As a matter of fact, most EP device will
only ask for 4MB MMIO window size.)

I will put this in the commit message.

thanks.
> 
> >  > -- 
> > > 2.6.4
> > > 
> 
>