search for: debug_gc

...ze/customize_main.ml b/customize/customize_main.ml index fa55c90..03c97e4 100644 --- a/customize/customize_main.ml +++ b/customize/customize_main.ml @@ -39,7 +39,6 @@ let main () = | s -> attach_format := Some s in let attach_disk s = attach := (!attach_format, s) :: !attach in - let debug_gc = ref false in let domain = ref None in let dryrun = ref false in let files = ref [] in @@ -79,7 +78,7 @@ let main () = "format" ^ " " ^ s_"Set attach disk format"; "-c", Arg.Set_string libvirt...

Nikita, Please take a look at the attached patch. I have rewritten it a little, and only lightly tested it. Rich.

...-git a/v2v/v2v.ml b/v2v/v2v.ml index b3dfa07..c20cbf0 100644 --- a/v2v/v2v.ml +++ b/v2v/v2v.ml @@ -144,40 +144,7 @@ let create_overlays src_disks = ov_virtual_size = vsize; ov_source = source } ) src_disks -let rec main () = - (* Handle the command line. *) - let input, output, - debug_gc, debug_overlays, do_copy, network_map, no_trim, - output_alloc, output_format, output_name, print_source, root_choice = - Cmdline.parse_cmdline () in - - (* Print the version, easier than asking users to tell us. *) - if verbose () then - printf "%s: %s %s (%s)\n%!" - prog...

...letions(-) diff --git a/v2v/v2v.ml b/v2v/v2v.ml index 9cb4a27..a2cf249 100644 --- a/v2v/v2v.ml +++ b/v2v/v2v.ml @@ -43,21 +43,7 @@ let print_mpstat chan { mp_dev = dev; mp_path = path; let () = Random.self_init () -let rec main () = - (* Handle the command line. *) - let input, output, - debug_gc, debug_overlays, do_copy, network_map, no_trim, - output_alloc, output_format, output_name, print_source, root_choice = - Cmdline.parse_cmdline () in - - (* Print the version, easier than asking users to tell us. *) - if verbose () then - printf "%s: %s %s (%s)\n%!" - prog...

...nsertions(+), 32 deletions(-) diff --git a/sysprep/main.ml b/sysprep/main.ml index d06b0d2..7d3ce5e 100644 --- a/sysprep/main.ml +++ b/sysprep/main.ml @@ -29,7 +29,7 @@ let () = Sysprep_operation.bake () (* Command line argument parsing. *) let prog = Filename.basename Sys.executable_name -let debug_gc, operations, g, selinux_relabel = +let debug_gc, operations, g, selinux_relabel, show_log = let debug_gc = ref false in let domain = ref None in let dryrun = ref false in @@ -40,6 +40,7 @@ let debug_gc, operations, g, selinux_relabel = let selinux_relabel = ref `Auto in let trace = r...

..._mode <> `Not_set && in_place then + error (f_"-o and --in-place cannot be used at the same time"); let output = match output_mode with | `Glance -> @@ -385,6 +390,6 @@ read the man page virt-v2v(1). vmtype output_alloc in input, output, - debug_gc, debug_overlays, do_copy, network_map, no_trim, + debug_gc, debug_overlays, do_copy, in_place, network_map, no_trim, output_alloc, output_format, output_name, print_source, root_choice diff --git a/v2v/v2v.ml b/v2v/v2v.ml index 1228316..88bbbaa 100644 --- a/v2v/v2v.ml +++ b/v2v/v2v.ml @@ -86...

...anged, 3 insertions(+), 5 deletions(-) diff --git a/v2v/v2v.ml b/v2v/v2v.ml index 7c47ea0..9cb4a27 100644 --- a/v2v/v2v.ml +++ b/v2v/v2v.ml @@ -55,6 +55,9 @@ let rec main () = printf "%s: %s %s (%s)\n%!" prog Config.package_name Config.package_version Config.host_cpu; + if debug_gc then + at_exit (fun () -> Gc.compact()); + message (f_"Opening the source %s") input#as_options; let source = input#source () in @@ -63,8 +66,6 @@ let rec main () = printf (f_"Source guest information (--print-source option):\n"); printf "\n";...

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.

...rt of this patch series? > > I think so. > > The goal of the refactoring part of the series was to reduce the amount > of detail in main(), and only leave coarse steps, making it easy to see > the whole scenario. > > Originally every exit path had a clause > > if debug_gc then > Gc.compact () > > There were two of them, and I was about to add another one. OK - I see, although only every non-error exit path. But adding an atexit handler ought to be safe, since the garbage collector should never be in an inconsistent state, and also the GC should never...

Hi All, The first time i noticed this problem is when i enabled some debug prints in memory.c DEBUG_GC DEBUG_ADJUST_HEAP DEBUG_RELEASE_MEM etc.. These have some prints for debugging .Base R package installation completes well and good. The problem comes when some of the Recommended package installation happens. The following Packages fail "KernSmooth,lattice,rpart,cluster,survival etc..."...

On Tue, Aug 11, 2015 at 08:00:20PM +0300, Roman Kagan wrote: > debub_gc (coming from the command line) indicates that gc should be > forced on program exit. Instead of sticking it on every exit path, > register it as an at_exit hook once. Was this change necessary as part of this patch series? The --debug-gc option is used across most of the virt-* tools in the internal tests, and

...iff --git a/sysprep/main.ml b/sysprep/main.ml index c1ce3c7..9431e88 100644 --- a/sysprep/main.ml +++ b/sysprep/main.ml @@ -31,6 +31,8 @@ let () = Sysprep_operation.bake () (* Command line argument parsing. *) let prog = Filename.basename Sys.executable_name +let () = Random.self_init () + let debug_gc, operations, g, selinux_relabel, quiet, mount_opts = let debug_gc = ref false in let domain = ref None in diff --git a/sysprep/script4.sh b/sysprep/script4.sh new file mode 100755 index 0000000..fe377e0 --- /dev/null +++ b/sysprep/script4.sh @@ -0,0 +1,21 @@ +#!/bin/bash - +# libguestfs virt-...

...ode, and changing that means having fewer, simpler patches in this series. And I think the in-place/copying stuff can be made type-safe as above. Rich. ---------------------------------------------------------------------- let main () = (* Handle the command line. *) let input, output, debug_gc, debug_overlays, do_copy, in_place, network_map, no_trim, output_alloc, output_format, output_name, print_source, root_choice = Cmdline.parse_cmdline () in (* Print the version, easier than asking users to tell us. *) if verbose () then printf "%s: %s %s (%s)\n%!" p...

On Friday 10 January 2014 10:09:19 Richard W.M. Jones wrote: > On Thu, Jan 09, 2014 at 03:45:54PM +0000, Richard W.M. Jones wrote: > > On Thu, Jan 09, 2014 at 04:21:10PM +0100, Pino Toscano wrote: > > > + and set_operations op_string = > > > + let currentopset = > > > + match (!operations) with > > > > No need for parentheses around

...| 21 ++++++++++++++++ sysprep/virt-sysprep.pod | 57 +++++++++++++++++++++++++++++++++++++------ 4 files changed, 131 insertions(+), 7 deletions(-) diff --git a/sysprep/main.ml b/sysprep/main.ml index 689a394..49750a9 100644 --- a/sysprep/main.ml +++ b/sysprep/main.ml @@ -87,6 +87,40 @@ let debug_gc, operations, g, selinux_relabel, quiet, mount_opts = exit 1 ) Sysprep_operation.empty_set ops in operations := Some opset + and set_operations op_string = + let currentopset = + match !operations with + | Some x -> x + | None -> Sysprep_operation.empty_...

This series is a second attempt to add a mode of virt-v2v operation where it leaves the config and disk image conversion, rollback on errors, registering with the destination hypervisor, etc. to a third-party toolset, and performs only tuning of the guest OS to run in the KVM-based hypervisor. The first 14 patches are just refactoring and rearrangement of the code, factoring the implementation

On Tue, Aug 11, 2015 at 08:00:34PM +0300, Roman Kagan wrote: > + let overlays = > + if not in_place then create_overlays source.s_disks > + else [] in > + let targets = > + if not in_place then init_targets overlays source output output_format > + else [] in This doesn't solve the problem I raised before which is that overlays and targets should never be empty

Version 1 of the patch, for discussion, but not to be applied. It does work, but needs a lot more testing. This removes all the psychopathic gettextify cruft, and replaces it with a 99 line Makefile.am. A large win, I think. The third patch implements gettext support in the OCaml tools. The fourth patch is just for illustration. It shows the consequent changes to libguestfs.pot and the po

...k once. > > Was this change necessary as part of this patch series? I think so. The goal of the refactoring part of the series was to reduce the amount of detail in main(), and only leave coarse steps, making it easy to see the whole scenario. Originally every exit path had a clause if debug_gc then Gc.compact () There were two of them, and I was about to add another one. So I went ahead and replaced them all with two lines at the beginning, setting up an at_exit hook. > The --debug-gc option is used across most of the virt-* tools in the > internal tests, and those tests sho...

...ink so. > > > > The goal of the refactoring part of the series was to reduce the amount > > of detail in main(), and only leave coarse steps, making it easy to see > > the whole scenario. > > > > Originally every exit path had a clause > > > > if debug_gc then > > Gc.compact () > > > > There were two of them, and I was about to add another one. > > OK - I see, although only every non-error exit path. But adding an > atexit handler ought to be safe, since the garbage collector should > never be in an inconsistent...