search for: hypervisor_grant_table_op

This is a cherry pick of two patches that add support for guest paged out frames in the XCP 2.6.32 dom0 patch queue. First patch propagates the ENOENT returned by the hypervisor in the case of a paged out page, all the way up the call chain to the MMAPBATCH_V2 ioctl. The ioctl is mainly used to harvest those return values and retry. The second patch adds retry loops to all backend grant

Patrick, there following patches fix xenpaging for me. Granttable handling is incomplete. If a page is gone, a GNTST_eagain should be returned to the caller to inidcate the hypercall has to be retried after a while, until the page is available again. Please review. Olaf _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com

In dispatch_rw_block_io after a call to HYPERVISOR_grant_table_op, there is the following code which calls fast_flush_area and breaks out of the loop early if one of the handles returned from HYPERVISOR_grant_table_op is negative: for (i = 0; i < nseg; i++) { if (unlikely(map[i].handle < 0)) { DPRINTK("invalid buffer -- could not remap it\n&quot...

...d!\n", (int)op.status); - ret = -EINVAL; - } - - return ret; -} - -static void unmap_frontend_page(blkif_t *blkif) -{ - struct gnttab_unmap_grant_ref op; - - gnttab_set_unmap_op(&op, (unsigned long)blkif->blk_ring_area->addr, - GNTMAP_host_map, blkif->shmem_handle); - - if (HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref, &op, 1)) - BUG(); -} - -int blkif_map(blkif_t *blkif, unsigned long shared_page, unsigned int evtchn) +int blkif_map(blkif_t *blkif, grant_ref_t ring_ref, evtchn_port_t evtchn) { + struct vm_struct *area; int err; /* Already connected through? */ if (blkif-&g...

This changeset: changeset: 15606:17820aa1b2c0 user: Keir Fraser <keir.fraser@citrix.com> date: Fri Dec 14 11:58:37 2007 +0000 summary: hvm: Remove duplicate assignment of __HYPERVISOR_grant_table_op in Has broken our HVM PV drivers. Previously we were using: 520 if ((HYPERVISOR_grant_table_op(GNTTABOP_setup_table, pset, 1) != 0) || 521 (pset->status != 0)) { even on HVM. Now, this isn''t actually necessary (stuff works fine with the call nop''d out, p...

Hi, I am trying to export a memory page of dom0 to a domU in readonly mode. In dom0, I grant access permission to that domU. In domU, I did HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, map, res->nr_pages); All these functions returned successfully. But when I tried to access data in the mapped memory page, all data I got is 0. Is there any restriction that preventing dom0 pages from being export to other domains? If so, is there any way to relax th...

I''m having problems starting more than three domains. It sometimes works fine, but more often than not the 4th domain''s root block device times out and so the domU kernel panics as there''s no /dev/root: XENBUS: Timeout connecting to device: device/vbd/2057 (state 6) XENBUS: Timeout connecting to device: device/vif/0 (state 6) XENBUS: Timeout connecting to device:

...for (i = 0; i < nr_grefs; i++) { + op[i].flags = GNTMAP_host_map | GNTMAP_contains_pte, + op[i].ref = gnt_ref[i], + op[i].dom = dev->otherend_id, + op[i].host_addr = arbitrary_virt_to_machine(pte[i]).maddr; + }; if (HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, &op, 1)) BUG(); - if (op.status != GNTST_okay) { - free_vm_area(area); - kfree(node); - xenbus_dev_fatal(dev, op.status, - "mapping in shared page %d from domain %d"...

...for (i = 0; i < nr_grefs; i++) { + op[i].flags = GNTMAP_host_map | GNTMAP_contains_pte, + op[i].ref = gnt_ref[i], + op[i].dom = dev->otherend_id, + op[i].host_addr = arbitrary_virt_to_machine(pte[i]).maddr; + }; if (HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, &op, 1)) BUG(); - if (op.status != GNTST_okay) { - free_vm_area(area); - kfree(node); - xenbus_dev_fatal(dev, op.status, - "mapping in shared page %d from domain %d"...

...for (i = 0; i < nr_grefs; i++) { + op[i].flags = GNTMAP_host_map | GNTMAP_contains_pte, + op[i].ref = gnt_ref[i], + op[i].dom = dev->otherend_id, + op[i].host_addr = arbitrary_virt_to_machine(pte[i]).maddr; + }; if (HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, &op, 1)) BUG(); - if (op.status != GNTST_okay) { - free_vm_area(area); - kfree(node); - xenbus_dev_fatal(dev, op.status, - "mapping in shared page %d from domain %d"...

...RVISOR_sysctl + * All generic sub-operations, with the exception of: + * * XEN_SYSCTL_page_offline_op + * * XEN_SYSCTL_get_pmstat + * * XEN_SYSCTL_pm_op + * + * HYPERVISOR_hvm_op + * Exactly these sub-operations are supported: + * * HVMOP_set_param + * * HVMOP_get_param + * + * HYPERVISOR_grant_table_op + * All generic sub-operations + * + * HYPERVISOR_vcpu_op + * Exactly these sub-operations are supported: + * * VCPUOP_register_vcpu_info + * * VCPUOP_register_runstate_memory_area + */ + #define XEN_HYPERCALL_TAG 0XEA1 #define uint64_aligned_t uint64_t __attribute__((aligned(8)))...

This patch is a significant performance improvement for the m2p_override: about 6% using the gntdev device. Each m2p_add/remove_override call issues a MULTI_grant_table_op and a __flush_tlb_single if kmap_op != NULL. Batching all the calls together is a great performance benefit because it means issuing one hypercall total rather than two hypercall per page. If paravirt_lazy_mode is set

...ex b84e3ab..6d325bd 100644 --- a/drivers/xen/grant-table.c +++ b/drivers/xen/grant-table.c @@ -933,9 +933,6 @@ int gnttab_map_refs(struct gnttab_map_grant_ref *map_ops, struct page **pages, unsigned int count) { int i, ret; - bool lazy = false; - pte_t *pte; - unsigned long mfn; ret = HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, map_ops, count); if (ret) @@ -947,45 +944,7 @@ int gnttab_map_refs(struct gnttab_map_grant_ref *map_ops, gnttab_retry_eagain_gop(GNTTABOP_map_grant_ref, map_ops + i, &map_ops[i].status, __func__); - /* this is basically a nop on x86 */ - if (xen_feature(XE...

...ex b84e3ab..6d325bd 100644 --- a/drivers/xen/grant-table.c +++ b/drivers/xen/grant-table.c @@ -933,9 +933,6 @@ int gnttab_map_refs(struct gnttab_map_grant_ref *map_ops, struct page **pages, unsigned int count) { int i, ret; - bool lazy = false; - pte_t *pte; - unsigned long mfn; ret = HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, map_ops, count); if (ret) @@ -947,45 +944,7 @@ int gnttab_map_refs(struct gnttab_map_grant_ref *map_ops, gnttab_retry_eagain_gop(GNTTABOP_map_grant_ref, map_ops + i, &map_ops[i].status, __func__); - /* this is basically a nop on x86 */ - if (xen_feature(XE...

...void * shared_page; extern grant_entry_t * grant_table; void offer_page() { uint16_t flags; /* create the grant table*/ gnttab_setup_table_t setup_op; setup_op.dom ＝DOMID＿SELF； setup_op.nr_frames = 1; setup_op.frame_list = grant_table; HYPERVISOR_grant_table_op(GNTTABOP_setup_table, &setup_op, 1); /*offer the grant */ grant_table[0].domid=DOMID_FRIEND; grant_table[0].frame = shared_page >>12; flgas = GTF_permit_access & GTF_reading & GTF_writing; grant_table[0].flags= flags; } and it also assume that t...

..._first_gref; for (i = 0; i < buffer_num_pages; i++) { printk(KERN_INFO "Mounting GREF %d\n", *grefp); memset(&op, 0, sizeof(op)); op.host_addr = x->buffer_addr + i * PAGE_SIZE; op.flags = GNTMAP_host_map; op.ref = *grefp; op.dom = x->otherend_id; rc = HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, &op, 1); if (rc == -ENOSYS) { goto err_unmap; } if (op.status) { DPRINTK("error: grant table mapping failed\n"); goto err_unmap; } x->buffer_handles[i] = op.handle; grefp = (int *)(x->buffer_addr + i * PAG...

I was following an email chain in the xen mailing list on HVM hypercalls on grant tables and the email chain link is below http://lists.xensource.com/archives/html/xen-devel/2007-03/msg00688.html This link says that we need not required hypercalls to share memory b/w HVM''s. But as far as my understanding of sharing memory is concerned, I think the memory is shared using grant table

I''ve been working form the netfront.c in the testing tree and using SLES 10 RC1 for i386 on a SMP box. When I stress the network using iperf in a domU, domU acting as client on a gigabit network, I occasionally get a panic at the dev_kfree_skb_irq(skb); line. This is the same panic as reported in http://lists.xensource.com/archives/html/xen-devel/2006-05/msg00919.html The trace

pvSCSI backend driver Signed-off-by: Tomonari Horikoshi <t.horikoshi@jp.fujitsu.com> Signed-off-by: Jun Kamada <kama@jp.fujitsu.com> ----- Jun Kamada _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel

...out writing a custom channel from scratch, and I''d like to know what the best way to proceed is. I''ll likely need to add a new VIRQ like the TRACE VIRQ and use that for notifications. However, allocating and sharing a new page is trickier. I''ve read about grant tables and HYPERVISOR_grant_table_op, but trace.c simply calls share_xen_page_with_privileged_guests(). I''ve read about ring buffers and DEFINE_RING_TYPES, but trace.{h,c} has nothing to do with that macro. Is the "The Definitive Guide to the Xen Hypervisor" book still relevant? What would you recommend for my case...