search for: escaped_file

QEMU upstream has broken snapshot=on ... again. These two patches stop using it entirely. Instead we run 'qemu-img create' to create overlay disks as required. Note that the libvirt and UML backends were already doing this: The libvirt backend because <transient/> has never worked, and the UML backend was running uml_mkcow because the UML-equivalent syntax of snapshot=on was

These patches contain the beginnings of discard (a.k.a. trim or unmap) support. This will allow us to change virt-sparsify to work on disk images in-place (instead of using slow & inefficient copying). The approach used is to add an optional 'discard' parameter to add-drive. It has 3 possible settings: - 'disable' : the default, no discard is done - 'besteffort' :

...estfs_h *g, void *datav, const char *arg) goto cleanup0; } #endif + else if (drv->iface && STREQ(drv->iface, "vhost-user-scsi-pci")) { + /* SPDK */ + ADD_CMDLINE ("-chardev"); + ADD_CMDLINE_PRINTF ("socket,id=vhost,path=%s", escaped_file); + ADD_CMDLINE ("-device"); + ADD_CMDLINE ("vhost-user-scsi-pci,chardev=vhost"); + } else if (drv->iface) { ADD_CMDLINE ("-drive"); ADD_CMDLINE_PRINTF ("%s,if=%s", param, drv->iface); diff --git a/lib/launch-libvirt.c b/...

...a second -device + * parameter to connect this drive to the SCSI HBA, as is required by + * virtio-scsi. + */ static char * qemu_drive_param (guestfs_h *g, const struct drive *drv, size_t index) { - size_t i; - size_t len = 128; - const char *p; - char *r; + CLEANUP_FREE char *file = NULL, *escaped_file = NULL; + size_t i, len; const char *iface; + char *p, *ret; + + /* Make the file= parameter. */ + switch (drv->protocol) { + case drive_protocol_file: + file = safe_strdup (g, drv->u.path); + break; + case drive_protocol_nbd: + if (STREQ (drv->u.nbd.exportname, "&qu...

Currently we have an OCaml library for generating the qemu command line (used only by ‘virt-v2v -o qemu’). However we also generate a qemu command line in ‘lib/launch-direct.c’, and we might in future need to generate a ‘-readconfig’-compatible configuration file if we want to go beyond 10,000 drives for scalability testing. Therefore this patch series reimplements the qemu command line code as

...INE ("-device"); - ADD_CMDLINE (VIRTIO_SCSI ",id=scsi"); - } + /* Create the virtio-scsi bus. */ + ADD_CMDLINE ("-device"); + ADD_CMDLINE (VIRTIO_SCSI ",id=scsi"); + /* Add drives */ ITER_DRIVES (g, i, drv) { CLEANUP_FREE char *file = NULL, *escaped_file = NULL, *param = NULL; @@ -529,10 +519,14 @@ launch_direct (guestfs_h *g, void *datav, const char *arg) } /* If there's an explicit 'iface', use it. Otherwise default to - * virtio-scsi if available. Otherwise default to virtio-blk. + * virtio-scsi. */ -...

This set of patches: - Adds new APIs to support discard in libguestfs. - Adds discard support to virt-format. - Adds discard support to virt-sysprep. - Implements virt-sparsify --in-place. This is now working, after fixing the rather stupid bug in fstrim. I've pushed the ones which were ACKed previously + the fstrim fix. Rich.

This still isn't working at the moment. See: http://marc.info/?t=139457409300003&r=1&w=2 This set of patches: - Adds new APIs to support discard in libguestfs. - Adds discard support to virt-format. - Adds discard support to virt-sysprep. - Implements virt-sparsify --in-place. Rich.

libguestfs has used double and triple underscores in identifiers. These aren't valid for global names in C++. (http://stackoverflow.com/a/228797) These large but completely mechanical patches change the illegal identifiers to legal ones. Rich.