[PATCH] fixup! qemu: backport patches to fix cves

[PATCH] fixup! qemu: backport patches to fix cves

[kai.kang]

From: Kai Kang <kai.kang@windriver.com>

Signed-off-by: Kai Kang <kai.kang@windriver.com>
---
 meta/recipes-devtools/qemu/qemu/0015-fix-CVE-2018-20124.patch | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/meta/recipes-devtools/qemu/qemu/0015-fix-CVE-2018-20124.patch b/meta/recipes-devtools/qemu/qemu/0015-fix-CVE-2018-20124.patch
index ad846958a7..985b819409 100644
--- a/meta/recipes-devtools/qemu/qemu/0015-fix-CVE-2018-20124.patch
+++ b/meta/recipes-devtools/qemu/qemu/0015-fix-CVE-2018-20124.patch
@@ -38,7 +38,7 @@ index d7a4bbd9..7f8028f8 100644
 -        comp_handler(IBV_WC_GENERAL_ERR, VENDOR_ERR_NO_SGE, ctx);
 +    if (!num_sge || num_sge > MAX_SGE) {
 +        pr_dbg("invalid num_sge=%d\n", num_sge);
-+        comp_handler(IBV_WC_GENERAL_ERR, VENDOR_ERR_INV_NUM_SGE, ctx);
++        comp_handler(IBV_WC_GENERAL_ERR, VENDOR_ERR_NO_SGE, ctx);
          return;
      }
  
@@ -51,7 +51,7 @@ index d7a4bbd9..7f8028f8 100644
 -        comp_handler(IBV_WC_GENERAL_ERR, VENDOR_ERR_NO_SGE, ctx);
 +    if (!num_sge || num_sge > MAX_SGE) {
 +        pr_dbg("invalid num_sge=%d\n", num_sge);
-+        comp_handler(IBV_WC_GENERAL_ERR, VENDOR_ERR_INV_NUM_SGE, ctx);
++        comp_handler(IBV_WC_GENERAL_ERR, VENDOR_ERR_NO_SGE, ctx);
          return;
      }
  
-- 
2.20.0

Re: [PATCH] fixup! qemu: backport patches to fix cves

[Alexey Brodkin]

Hi Kai Kang,

> -----Original Message-----
> From: kai.kang@windriver.com <kai.kang@windriver.com>
> Sent: Wednesday, March 20, 2019 4:16 PM
> To: richard.purdie@linuxfoundation.org
> Cc: alexey.brodkin@synopsys.com; openembedded-core@lists.openembedded.org
> Subject: [PATCH] fixup! qemu: backport patches to fix cves

Just wondering if you were able to reproduce mentioned problem
on your side before fixing it?

Given Richard's input this problem was supposed to be caught
before your previous commit hit OE master branch but somehow it
slipped through the cracks...

Or otherwise my setup should be somehow very special if it only
happens for me.

-Alexey

Re: [PATCH] fixup! qemu: backport patches to fix cves

[Kang Kai]

On 2019/3/21 上午5:34, Alexey Brodkin wrote:
> Hi Kai Kang,
>
>> -----Original Message-----
>> From: kai.kang@windriver.com <kai.kang@windriver.com>
>> Sent: Wednesday, March 20, 2019 4:16 PM
>> To: richard.purdie@linuxfoundation.org
>> Cc: alexey.brodkin@synopsys.com; openembedded-core@lists.openembedded.org
>> Subject: [PATCH] fixup! qemu: backport patches to fix cves
> Just wondering if you were able to reproduce mentioned problem
> on your side before fixing it?
>
> Given Richard's input this problem was supposed to be caught
> before your previous commit hit OE master branch but somehow it
> slipped through the cracks...
>
> Or otherwise my setup should be somehow very special if it only
> happens for me.

I can reproduce the build failure with default configure:

$ . poky/oe-init-build-env
$ bitbake qemu-system-native

It should be covered by autobuilder. :(

And there is no error to build qemu and qemu-native  that 
hw/rdma/rdma_backend.c is not even compiled because:

* CONFIG_PCI is not set for qemu-native

* CONFIG_PVRDMA is not set for qemu


Regards,
Kai




>
> -Alexey
>
>

-- 
Kai Kang

Re: [PATCH] fixup! qemu: backport patches to fix cves

[Alexey Brodkin]

Hi Kai,

> -----Original Message-----
> From: Kang Kai <Kai.Kang@windriver.com>
> Sent: Thursday, March 21, 2019 6:27 AM
> To: Alexey Brodkin <alexey.brodkin@synopsys.com>
> Cc: openembedded-core@lists.openembedded.org; richard.purdie@linuxfoundation.org
> Subject: Re: [PATCH] fixup! qemu: backport patches to fix cves
> 
> On 2019/3/21 上午5:34, Alexey Brodkin wrote:
> > Hi Kai Kang,
> >
> >> -----Original Message-----
> >> From: kai.kang@windriver.com <kai.kang@windriver.com>
> >> Sent: Wednesday, March 20, 2019 4:16 PM
> >> To: richard.purdie@linuxfoundation.org
> >> Cc: alexey.brodkin@synopsys.com; openembedded-core@lists.openembedded.org
> >> Subject: [PATCH] fixup! qemu: backport patches to fix cves
> > Just wondering if you were able to reproduce mentioned problem
> > on your side before fixing it?
> >
> > Given Richard's input this problem was supposed to be caught
> > before your previous commit hit OE master branch but somehow it
> > slipped through the cracks...
> >
> > Or otherwise my setup should be somehow very special if it only
> > happens for me.
> 
> I can reproduce the build failure with default configure:
> 
> $ . poky/oe-init-build-env
> $ bitbake qemu-system-native
> 
> It should be covered by autobuilder. :(
> 
> And there is no error to build qemu and qemu-native  that
> hw/rdma/rdma_backend.c is not even compiled because:
> 
> * CONFIG_PCI is not set for qemu-native
> * CONFIG_PVRDMA is not set for qemu

Thanks for confirmation!

Richard, FWIW I hit this issue building "build-appliance-image"
which I use for smoke-testing as there're quite a few packages
being built.

-Alexey