DPDK patches and discussions
 help / color / mirror / Atom feed
From: Thomas Monjalon <thomas.monjalon@6wind.com>
To: "SMonderer@advaoptical.com" <SMonderer@advaoptical.com>
Cc: "dev@dpdk.org" <dev@dpdk.org>,
	"xen-devel@lists.xen.org" <xen-devel@lists.xen.org>
Subject: Re: [dpdk-dev] could not l2fwd in DOM0
Date: Mon, 09 Jun 2014 18:02:17 +0200	[thread overview]
Message-ID: <1959134.aMFyYmvJE8@xps13> (raw)
In-Reply-To: <1ED644BD7E0A5F4091CF203DAFB8E4CC01C7AB55@SHSMSX101.ccr.corp.intel.com>

Jijiang's patch has been applied on master branch:
	http://dpdk.org/browse/dpdk/commit/?id=6f0ce7b9cd490

2014-05-29 10:12, Liu, Jijiang:
> Ok, I will send a patch for fixing this.
> 
> -----Original Message-----
> From: Thomas Monjalon [mailto:thomas.monjalon@6wind.com]
> Sent: Thursday, May 29, 2014 5:59 PM
> To: Liu, Jijiang
> Cc: dev@dpdk.org; SMonderer@advaoptical.com; xen-devel@lists.xen.org
> Subject: Re: [dpdk-dev] could not l2fwd in DOM0
> 
> Jijiang, I'm not sure to understand.
> Could you send a proper patch for this issue?
> 
> Thanks
> 
> > In Linux kernel versions < 3.14 , the vstart is kernel virtual address,
> > the
> > API prototype below. int xen_create_contiguous_region(unsigned long
> > vstart,
> > unsigned int order, unsigned int address_bits)
> > 
> > In Linux kernel 3.14, the pstart is kernel physical address, the API
> > prototype below. int xen_create_contiguous_region(phys_addr_t pstart,
> > unsigned int order, unsigned int address_bits,
> > 
> > 				 dma_addr_t *dma_handle)
> > 
> > ,so the mm_data->block_info[i].vir_addr cannot pass to
> > xen_create_contiguous_region() directly for Linux kernel 3.14, the pstart
> > should be mm_data->block_info[i].pfn * PAGE_SIZE in the Dom0 Driver.
> > 
> > #include <linux/errno.h>
> > 
> >  #include <linux/vmalloc.h>
> >  #include <linux/mm.h>
> >  
> >  #include <xen/xen.h>
> >  #include <xen/page.h>
> > 
> > @@ -309,6 +310,7 @@ dom0_prepare_memsegs(struct memory_info* meminfo,
> > struct dom0_mm_data uint64_t pfn, vstart, vaddr;
> > 
> >         uint32_t i, num_block, size;
> >         int idx;
> > 
> > +       dma_addr_t dma_handle;
> > 
> >         /* Allocate 2M memory once */
> >         num_block = meminfo->size / 2;
> > 
> > @@ -344,7 +346,7 @@ dom0_prepare_memsegs(struct memory_info* meminfo,
> > struct dom0_mm_data * contiguous physical addresses, its maximum size is
> > 2M. */> 
> >                 if
> >                 (xen_create_contiguous_region(mm_data->block_info[i].pfn
> > 
> > * PAGE_SIZE, -                                   DOM0_CONTIG_NUM_ORDER, 0)
> > == 0) { +                                   DOM0_CONTIG_NUM_ORDER, 0,
> > + &dma_handle) == 0) {
> > 
> >                         mm_data->block_info[i].exchange_flag = 1;
> >                         mm_data->block_info[i].mfn =
> >                         
> >                                 pfn_to_mfn(mm_data->block_info[i].pfn);
> > 
> > Thanks
> > Frank Liu
> > 
> > -----Original Message-----
> > From: dev [mailto:dev-bounces@dpdk.org] On Behalf Of Samuel Monderer
> > Sent: Wednesday, May 07, 2014 8:44 PM
> > To: Liu, Jijiang
> > Cc: dev@dpdk.org
> > Subject: Re: [dpdk-dev] could not l2fwd in DOM0
> > 
> > > -----Original Message-----
> > > From: dev [mailto:dev-bounces@dpdk.org] On Behalf Of Liu, Jijiang
> > > Sent: Wednesday, May 07, 2014 7:36 AM
> > > To: dev@dpdk.org
> > > Subject: Re: [dpdk-dev] could not l2fwd in DOM0
> > > 
> > > Hi,
> > > 
> > > I have checked source codes of xen_create_contiguous_region function
> > > in kernel 3.14, and found the dma_handle cannot be NULL.
> > > 
> > > int xen_create_contiguous_region(phys_addr_t pstart, unsigned int order,
> > > 
> > >                                  unsigned int address_bits,
> > >                                  dma_addr_t *dma_handle) {
> > >         
> > >         unsigned long *in_frames = discontig_frames, out_frame;
> > >         unsigned long  flags;
> > >         int            success;
> > >         unsigned long vstart = (unsigned long)phys_to_virt(pstart);
> > >     
> > >     ...
> > >     
> > >         *dma_handle = virt_to_machine(vstart).maddr;
> > >         return success ? 0 : -ENOMEM;
> > > 
> > > }
> > > 
> > > Thanks
> > > Frank Liu
> > 
> > Thanks Frank,
> > 
> > I've changed the code as following but now the kernel module crashes.
> > 
> > diff --git a/lib/librte_eal/linuxapp/xen_dom0/dom0_mm_misc.c
> > b/lib/librte_eal/linuxapp/xen index 87fa3e6..8addc21 100644
> > --- a/lib/librte_eal/linuxapp/xen_dom0/dom0_mm_misc.c
> > +++ b/lib/librte_eal/linuxapp/xen_dom0/dom0_mm_misc.c
> > @@ -64,6 +64,7 @@
> > 
> >  #include <linux/errno.h>
> >  #include <linux/vmalloc.h>
> >  #include <linux/mm.h>
> >  
> >  #include <xen/xen.h>
> >  #include <xen/page.h>
> > 
> > @@ -309,6 +310,7 @@ dom0_prepare_memsegs(struct memory_info* meminfo,
> > struct dom0_mm_data uint64_t pfn, vstart, vaddr;
> > 
> >         uint32_t i, num_block, size;
> >         int idx;
> > 
> > +       dma_addr_t dma_handle;
> > 
> >         /* Allocate 2M memory once */
> >         num_block = meminfo->size / 2;
> > 
> > @@ -344,7 +346,7 @@ dom0_prepare_memsegs(struct memory_info* meminfo,
> > struct dom0_mm_data * contiguous physical addresses, its maximum size is
> > 2M. */> 
> >                 if
> > 
> > (xen_create_contiguous_region(mm_data->block_info[i].vir_addr,
> > -                                   DOM0_CONTIG_NUM_ORDER, 0) == 0) {
> > +                                   DOM0_CONTIG_NUM_ORDER, 0,
> > + &dma_handle) == 0) {
> > 
> >                         mm_data->block_info[i].exchange_flag = 1;
> >                         mm_data->block_info[i].mfn =
> >                         
> >                                 pfn_to_mfn(mm_data->block_info[i].pfn);
> > 
> > Samuel
> > 
> > > -----Original Message-----
> > > From: dev [mailto:dev-bounces@dpdk.org] On Behalf Of Samuel Monderer
> > > Sent: Wednesday, April 30, 2014 1:54 AM
> > > To: dev@dpdk.org<mailto:dev@dpdk.org>
> > > Cc: Shimon Zadok
> > > Subject: [dpdk-dev] could not l2fwd in DOM0
> > > 
> > > Hi,
> > > 
> > > First I've encountered a compiling problem when compiling for DOM0 due
> > > to prototype change of the function xen_create_contiguous_region I
> > > made the following changes:
> > > 
> > > diff --git a/lib/librte_eal/linuxapp/xen_dom0/dom0_mm_misc.c
> > > b/lib/librte_eal/linuxapp/xen index 87fa3e6..8addc21 100644
> > > --- a/lib/librte_eal/linuxapp/xen_dom0/dom0_mm_misc.c
> > > +++ b/lib/librte_eal/linuxapp/xen_dom0/dom0_mm_misc.c
> > > @@ -64,6 +64,7 @@
> > > 
> > >  #include <linux/errno.h>
> > >  #include <linux/vmalloc.h>
> > >  #include <linux/mm.h>
> > > 
> > > +//#include <linux/types.h>
> > > 
> > >  #include <xen/xen.h>
> > >  #include <xen/page.h>
> > > 
> > > @@ -309,6 +310,7 @@ dom0_prepare_memsegs(struct memory_info* meminfo,
> > > struct dom0_mm_data
> > > 
> > >         uint64_t pfn, vstart, vaddr;
> > >         uint32_t i, num_block, size;
> > >         int idx;
> > > 
> > > +       dma_addr_t *dma_handle = NULL;
> > > 
> > >         /* Allocate 2M memory once */
> > >         num_block = meminfo->size / 2; @@ -344,7 +346,7 @@
> > > 
> > > dom0_prepare_memsegs(struct memory_info* meminfo, struct dom0_mm_data
> > > 
> > >                  * contiguous physical addresses, its maximum size is
> > >                  2M.
> > >                  */
> > >                 
> > >                 if
> > >                 (xen_create_contiguous_region(mm_data->block_info[i].vir
> > >                 _
> > >                 addr,
> > > 
> > > -                                   DOM0_CONTIG_NUM_ORDER, 0) == 0) {
> > > +                                   DOM0_CONTIG_NUM_ORDER, 0,
> > > + dma_handle) == 0) {
> > > 
> > >                         mm_data->block_info[i].exchange_flag = 1;
> > >                         mm_data->block_info[i].mfn =
> > > 
> > > pfn_to_mfn(mm_data->block_info[i].pfn);
> > > 
> > > After that I tried to run l2fwd example and got a segmentation fault
> > > 
> > > root@Smart:~/samuelm/dpdk/dpdk# modprobe uio
> > > root@Smart:~/samuelm/dpdk/dpdk# insmod ./x86_64-default-linuxapp-
> > > gcc/kmod/igb_uio.ko root@Smart:~/samuelm/dpdk/dpdk# insmod
> > > ./x86_64-default-linuxapp- gcc/kmod/rte_dom0_mm.ko
> > > root@Smart:~/samuelm/dpdk/dpdk# cd examples/l2fwd/build/
> > > root@Smart:~/samuelm/dpdk/dpdk/examples/l2fwd/build# echo 2048 >
> > > /sys/kernel/mm/dom0-mm/memsize-mB/memsize
> > > root@Smart:~/samuelm/dpdk/dpdk/examples/l2fwd/build# ./l2fwd -c 3 -n 4
> > > --xen-dom0 -- -q 1 -p 3
> > > EAL: Detected lcore 0 as core 0 on socket 0
> > > EAL: Detected lcore 1 as core 0 on socket 0
> > > EAL: Setting up memory...
> > > Segmentation fault
> > > root@Smart:~/samuelm/dpdk/dpdk/examples/l2fwd/build#
> > > 
> > > Has anyone already encountered this problem?
> > > 
> > > Samuelm

  reply	other threads:[~2014-06-09 16:02 UTC|newest]

Thread overview: 7+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-05-17  7:48 Liu, Jijiang
2014-05-29  9:58 ` Thomas Monjalon
2014-05-29 10:12   ` Liu, Jijiang
2014-06-09 16:02     ` Thomas Monjalon [this message]
  -- strict thread matches above, loose matches on Subject: below --
2014-04-29 17:53 Samuel Monderer
     [not found] ` <CAD16F236028A64DBBC0158B1636EA4510E8816E@SHSMSX104.ccr.corp.intel.com>
2014-05-07  4:35   ` Liu, Jijiang
     [not found]     ` <9f34498d15cd400eb2db76212778c038@MUC-SRV-MBX2.advaoptical.com>
2014-05-07 12:44       ` Samuel Monderer

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=1959134.aMFyYmvJE8@xps13 \
    --to=thomas.monjalon@6wind.com \
    --cc=SMonderer@advaoptical.com \
    --cc=dev@dpdk.org \
    --cc=xen-devel@lists.xen.org \
    /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).