test suite reviews and discussions
 help / color / mirror / Atom feed
From: "Tu, LijuanX A" <lijuanx.a.tu@intel.com>
To: "Liu, Yong" <yong.liu@intel.com>, "dts@dpdk.org" <dts@dpdk.org>
Subject: Re: [dts] [PATCH V1] framework: add kill qemu script. If vm not quit	normally, VF will not released. and this will cause next testsuite failed.
Date: Thu, 12 Jan 2017 03:16:38 +0000	[thread overview]
Message-ID: <38D041F150D4184C8114E499040E62343D1DFAD4@shsmsx102.ccr.corp.intel.com> (raw)
In-Reply-To: <86228AFD5BCD8E4EBFD2B90117B5E81E62D42F71@SHSMSX103.ccr.corp.intel.com>

Hi yong
	I have got your comments, will sent v2 as soon as possible.

> Do not known why need to enable autoneg on PF device.
[lijuan] I will remove this ,it's not related with the function, though it's know issue, refer to DPDK-1754

 > -----Original Message-----
 > From: Liu, Yong
 > Sent: Thursday, January 12, 2017 9:43 AM
 > To: Tu, LijuanX A; dts@dpdk.org
 > Cc: Tu, LijuanX A
 > Subject: RE: [dts] [PATCH V1] framework: add kill qemu script. If vm not quit
 > normally, VF will not released. and this will cause next testsuite failed.
 > 
 > Lijuan,
 > Some comments below.  Please make the email subject more compact and add
 > detail description in your git comment.
 > 
 > > -----Original Message-----
 > > From: dts [mailto:dts-bounces@dpdk.org] On Behalf Of Lijuan Tu
 > > Sent: Monday, January 09, 2017 2:05 PM
 > > To: dts@dpdk.org
 > > Cc: Tu, LijuanX A <lijuanx.a.tu@intel.com>
 > > Subject: [dts] [PATCH V1] framework: add kill qemu script. If vm not
 > > quit normally, VF will not released. and this will cause next testsuite
 > failed.
 > >
 > > Signed-off-by: Lijuan Tu <lijuanx.a.tu@intel.com>
 > > ---
 > >  framework/dut.py       | 24 ++++++++++++++++++++++++
 > >  framework/qemu_kvm.py  | 30 ++++++++++++++++++++++++++++++
 > > framework/test_case.py |  6 ++++++
 > >  3 files changed, 60 insertions(+)
 > >
 > > diff --git a/framework/dut.py b/framework/dut.py index
 > > 1d4a383..9500107
 > > 100644
 > > --- a/framework/dut.py
 > > +++ b/framework/dut.py
 > > @@ -73,6 +73,8 @@ class Dut(Crb):
 > >          self.conf = PortConf()
 > >          self.ports_map = []
 > >          self.virt_pool = None
 > > +        # hypervisor pid list, used for cleanup
 > > +        self.virt_pids = []
 > >
 > >      def init_host_session(self):
 > >          if self.host_init_flag:
 > > @@ -816,6 +818,17 @@ class Dut(Crb):
 > >          pci = self.ports_info[port_id]['pci']
 > >          self.virt_pool.del_vf_on_pf(pf_pci=pci, vflist=vflist)
 > >
 > > +    def destroy_all_sriov_vfs(self):
 > > +
 > > +        if self.ports_info == None:
 > > +            return
 > > +        for port_id in (0,len(self.ports_info)-1):
 > Not satisfy with pep8 rule, please check first.
 > Parameter port_id in destroy_sriov_vfs_by_port mean port index, the function
 > here only destroy devices of the first and last port.
 > You need use code like "range(len(self.ports_info))" to loop on all ports.
 > 
 > > +            print port_id
 > 
 > Need remove this debug code.
 > 
 > > +            self.destroy_sriov_vfs_by_port(port_id)
 > > +            # know issue: nic up
 > I can't understand this comment, please add more description.
 > 
 > > +            intf = self.ports_info[port_id]['intf']
 > > +            self.send_expect("ethtool -s %s autoneg on" % intf, "# ")
 > > +
 > 
 > Do not known why need to enable autoneg on PF device.
[lijuan] I will remove this ,it's not related with the function, though it's know issue, refer to DPDK-1754
 > 
 > >      def get_vm_core_list(self):
 > >          return VMCORELIST[self.crb['VM CoreList']]
 > >
 > > @@ -966,6 +979,17 @@ class Dut(Crb):
 > >          if self.host_init_flag:
 > >              self.host_session.close()
 > >
 > > +    def virt_exit(self):
 > > +        """
 > > +        Stop all unstopped hypervisors process
 > > +        """
 > > +        # try to kill all hypervisor process
 > > +        for pid in self.virt_pids:
 > > +            print "kill %d" % pid
 > 
 > Please remove debug code here.
 > 
 > > +            self.send_expect("kill -s SIGTERM %d" % pid, "# ",
 > alt_session=True)
 > > +            time.sleep(3)
 > > +        self.virt_pids = []
 > > +
 > >      def crb_exit(self):
 > >          """
 > >          Recover all resource before crb exit diff --git
 > > a/framework/qemu_kvm.py b/framework/qemu_kvm.py index e9d29cc..79e8417
 > > 100644
 > > --- a/framework/qemu_kvm.py
 > > +++ b/framework/qemu_kvm.py
 > > @@ -105,6 +105,7 @@ class QEMUKvm(VirtBase):
 > >      def set_vm_default(self):
 > >          self.set_vm_name(self.vm_name)
 > >          self.set_vm_enable_kvm()
 > > +        self.set_vm_pid_file()
 > >          self.set_vm_qga()
 > >          self.set_vm_daemon()
 > >          self.set_vm_monitor()
 > > @@ -244,6 +245,25 @@ class QEMUKvm(VirtBase):
 > >              enable_kvm_boot_line = '-enable-kvm'
 > >              self.__add_boot_line(enable_kvm_boot_line)
 > >
 > > +    def set_vm_pid_file(self):
 > > +        """
 > > +        Set VM pidfile option for manage qemu process
 > > +        """
 > > +        self.__pid_file = '/tmp/.%s.pid' % self.vm_name
 > > +        index = self.find_option_index('pid_file')
 > > +        if index:
 > > +            self.params[index] = {'pid_file': [{'name': '%s' %
 > self.__pid_file}]}
 > > +        else:
 > > +            self.params.append({'pid_file': [{'name': '%s' %
 > > + self.__pid_file}]})
 > > +
 > > +    def add_vm_pid_file(self, **options):
 > > +        """
 > > +        'name' : '/tmp/.qemu_vm0.pid'
 > > +        """
 > > +        if 'name' in options.keys():
 > > +            self.__add_boot_line('-pidfile %s' % options['name'])
 > > +
 > > +
 > >      def set_vm_name(self, vm_name):
 > >          """
 > >          Set VM name.
 > > @@ -1155,6 +1175,16 @@ class QEMUKvm(VirtBase):
 > >          else:
 > >              self.vm_status = ST_UNKNOWN
 > >
 > > +        info = self.host_session.send_expect('cat %s' % self.__pid_file,
 > "# ")
 > > +        try:
 > > +            pid = int(info)
 > > +            # save pid into dut structure
 > > +            self.host_dut.virt_pids.append(pid)
 > > +        except:
 > > +            self.host_logger.info("Failed to capture pid!!!")
 > > +
 > > +
 > > +
 > >      def __strip_guest_pci(self):
 > >          """
 > >          Strip all pci-passthrough device information, based on qemu
 > > monitor diff -- git a/framework/test_case.py b/framework/test_case.py
 > > index
 > > 85fdb70..270f7b9 100644
 > > --- a/framework/test_case.py
 > > +++ b/framework/test_case.py
 > > @@ -335,6 +335,12 @@ class TestCase(object):
 > >              dutobj.kill_all()
 > >          self.tester.kill_all()
 > >
 > > +        for dutobj in self.duts:
 > > +            dutobj.virt_exit()
 > > +            # destroy all vfs
 > > +            dutobj.destroy_all_sriov_vfs()
 > > +
 > > +
 > >      def wirespeed(self, nic, frame_size, num_ports):
 > >          """
 > >          Calculate bit rate. It is depended for NICs
 > > --
 > > 1.9.3

      reply	other threads:[~2017-01-12  3:16 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-01-09  6:05 Lijuan Tu
2017-01-12  1:42 ` Liu, Yong
2017-01-12  3:16   ` Tu, LijuanX A [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=38D041F150D4184C8114E499040E62343D1DFAD4@shsmsx102.ccr.corp.intel.com \
    --to=lijuanx.a.tu@intel.com \
    --cc=dts@dpdk.org \
    --cc=yong.liu@intel.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).