Date: Mon, 4 Sep 2017 19:25:14 +0530 From: Aijaz Baig <aijazbaig1@gmail.com> To: Julian Elischer <julian@freebsd.org> Cc: Gary Palmer <gpalmer@freebsd.org>, Bakul Shah <bakul@bitblocks.com>, Kirk McKusick <mckusick@mckusick.com>, freebsd-fs@freebsd.org Subject: Re: Tips on remote debugging for filesystem code Message-ID: <CAHB2L%2Be7UogO9fg6G_SnZDcgQnCd=z=Um=JqgXtLMeWnsGZ2yQ@mail.gmail.com> In-Reply-To: <e1fca65c-4468-295e-277f-e2c4b916978d@freebsd.org> References: <CAHB2L%2BcbBTfT%2Bpvg=c_39EKBSoLAo-_2KdmCcYgURQK5Gio4Xg@mail.gmail.com> <201708250251.v7P2pwcl029687@chez.mckusick.com> <CAHB2L%2BcrbMGvwAB7zx%2BnaLC5DTbOy7RWbxyFyPCGAZmaLoQ5fg@mail.gmail.com> <20170827195428.49400124AE9F@mail.bitblocks.com> <CAHB2L%2BfWd2WBU9GKNK8XGmmwvDTSFRUGChfhxozQ-k8Lj-VPBw@mail.gmail.com> <20170828132647.GA60563@in-addr.com> <CAHB2L%2BcicR7qSTGyhhhbx4eFtMn6P9snEzn5yac9WzXPL=pDhw@mail.gmail.com> <CAHB2L%2Bcrdda9fDdRnSvrZ46Pc8vbeJTdFAz390537Nnu9RPdDQ@mail.gmail.com> <e1fca65c-4468-295e-277f-e2c4b916978d@freebsd.org>
next in thread | previous in thread | raw e-mail | index | archive | help
Hi Julian Yes it just hangs. Yes it's not like it just hangs every time. It's more intermittent. This time I was stepping through 'kern_openat', really slowly as I am new to this, checking out the structures and the likes and finally this hit when I was within vn_open_cred. So yes the total time between attaching to the stub within the VM and this happening was close to 3 hours but nonetheless I do face this eventually is there something I should watch out for, I mean is there something time bound in some way for any of this code? Is this the reason why the VM becomes unresponsive if I step really slowly through the stack trace? On Mon, Sep 4, 2017 at 10:25 AM, Julian Elischer <julian@freebsd.org> wrote: > On 2/9/17 8:40 pm, Aijaz Baig wrote: > > I was finally able to deploy a FreeBSD-Current VM on Bhyve but even with > that setup I face a similar issue > > For instance I put a breakpoint on 'ffs_read' and let the debugged VM > continue but the control is neither returned to (k)gdb nor does the VM > actually run. It appears hung. > > are you sure it's not stopped at the breakpoint in ddb on the console? > > > Is there something within the ffs/ufs code that I need to be aware of that > it doesn't facilitate this sort of debugging? Or is it because Bhyve is > being run on FreeBSD which is itself run on another hypervisor? Will that > affect this? > > On Mon, Aug 28, 2017 at 7:03 PM, Aijaz Baig <aijazbaig1@gmail.com> wrote: > >> HI Gary >> >> Just checked that. Indeed that was the case. In addition I was running >> FreeBSD on top of vmware hypervisor and had not allowed hardware >> acceleration features pass through to the FreeBSD VM. It is now working! >> Need to explore more on getting it to connect to GDB >> >> Perhaps you'll see a thread soon for it (Has anything worked perfectly >> for the first time ever?? ;) )! >> >> On Mon, Aug 28, 2017 at 6:56 PM, Gary Palmer <gpalmer@freebsd.org> wrote: >> >>> On Mon, Aug 28, 2017 at 01:12:09PM +0530, Aijaz Baig wrote: >>> > Thanks a lot bakul. Will certain look into this >>> > >>> > Nevertheless I got hold of a PowerEdge 320 system which DOES have a >>> POPCNT >>> > instruction which shows support for V-x support via EPT but I still >>> get the >>> > same error there >>> > >>> > sh /usr/share/examples/bhyve/vmrun.sh -c 1 -m 2048M -t tap0 -d >>> guest.img -i >>> > -I FreeBSD-10.3-RELEASE-amd64-bootonly.iso fbsd10 >>> > Launching virtual machine "fbsd10" ... >>> > vm_create: Device not configured >>> > >>> > dmesg: >>> > vmx_init: processor does not support VMX operation >>> > >>> > cat /var/run/dmesg.boot | grep -i popcnt >>> > >>> > Features2=0xffba2203<SSE3,PCLMULQDQ,SSSE3,CX16,PCID,SSE4.1,S >>> SE4.2,x2APIC,POPCNT,TSCDLT,AESNI,XSAVE,OSXSAVE,AVX,F16C,RDRAND,HV> >>> > >>> > Features2=0xffba2203<SSE3,PCLMULQDQ,SSSE3,CX16,PCID,SSE4.1,S >>> SE4.2,x2APIC,POPCNT,TSCDLT,AESNI,XSAVE,OSXSAVE,AVX,F16C,RDRAND,HV> >>> > >>> > Any ideas? >>> >>> Hi, >>> >>> Check the BIOS for virtualisation or other related options. Some BIOSes >>> have >>> the ability to turn those features off, even if the processor normally >>> supports them >>> >>> Regards, >>> >>> Gary >>> >> >> >> >> -- >> >> Best Regards, >> Aijaz Baig >> > > > > -- > > Best Regards, > Aijaz Baig > > > -- Best Regards, Aijaz Baig
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CAHB2L%2Be7UogO9fg6G_SnZDcgQnCd=z=Um=JqgXtLMeWnsGZ2yQ>