From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:33877) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1YxQ1u-00020Z-I4 for qemu-devel@nongnu.org; Tue, 26 May 2015 21:21:12 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1YxQ1r-0002en-A1 for qemu-devel@nongnu.org; Tue, 26 May 2015 21:21:10 -0400 Received: from mail-oi0-x231.google.com ([2607:f8b0:4003:c06::231]:32861) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1YxQ1q-0002eY-RX for qemu-devel@nongnu.org; Tue, 26 May 2015 21:21:07 -0400 Received: by oiww2 with SMTP id w2so91668603oiw.0 for ; Tue, 26 May 2015 18:21:05 -0700 (PDT) Sender: fluxion Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable From: Michael Roth References: <1431327525-25625-1-git-send-email-justin@quarantainenet.nl> <1431327525-25625-3-git-send-email-justin@quarantainenet.nl> <5562FC52.80908@parallels.com> In-Reply-To: <5562FC52.80908@parallels.com> Message-ID: <20150527012054.5594.7649@loki> Date: Tue, 26 May 2015 20:20:54 -0500 Subject: Re: [Qemu-devel] [PATCH v4 2/2] qga/qmp_guest_fstrim: Return per path fstrim result List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Olga Krishtal , Justin Ossevoort , qemu-devel@nongnu.org Quoting Olga Krishtal (2015-05-25 05:41:22) > On 11/05/15 09:58, Justin Ossevoort wrote: > = > The current guest-fstrim support only returns an error if some > mountpoint was unable to be trimmed, skipping any possible additional > mountpoints. The result of the TRIM operation itself is also discarde= d. > = > This change returns a per mountpoint result of the TRIM operation. If= an > error occurs on some mountpoints that error is returned and the > guest-fstrim continue with any additional mountpoints. > = > The returned values for errors, minimum and trimmed are dependant on = the > filesystem, storage stacks and kernel version. > = > Signed-off-by: Justin Ossevoort > --- > qga/commands-posix.c | 54 ++++++++++++++++++++++++++++++++++++++----= ---------- > qga/commands-win32.c | 4 +++- > qga/qapi-schema.json | 32 ++++++++++++++++++++++++++++--- > 3 files changed, 72 insertions(+), 18 deletions(-) > = > diff --git a/qga/commands-posix.c b/qga/commands-posix.c > index 4449628..ec0d69e 100644 > --- a/qga/commands-posix.c > +++ b/qga/commands-posix.c > @@ -1325,8 +1325,12 @@ static void guest_fsfreeze_cleanup(void) > /* > * Walk list of mounted file systems in the guest, and trim them. > */ > -void qmp_guest_fstrim(bool has_minimum, int64_t minimum, Error **err= p) > +GuestFilesystemTrimResponse * > +qmp_guest_fstrim(bool has_minimum, int64_t minimum, Error **errp) > { > + GuestFilesystemTrimResponse *response; > + GuestFilesystemTrimResultList *list; > + GuestFilesystemTrimResult *result; > int ret =3D 0; > FsMountList mounts; > struct FsMount *mount; > @@ -1340,39 +1344,59 @@ void qmp_guest_fstrim(bool has_minimum, int64= _t minimum, Error **errp) > build_fs_mount_list(&mounts, &local_err); > if (local_err) { > error_propagate(errp, local_err); > - return; > + return NULL; > } > = > + response =3D g_malloc0(sizeof(*response)); > + > QTAILQ_FOREACH(mount, &mounts, next) { > + result =3D g_malloc0(sizeof(*result)); > + result->path =3D g_strdup(mount->dirname); > + > + list =3D g_malloc0(sizeof(*list)); > + list->value =3D result; > + list->next =3D response->paths; > + response->paths =3D list; > + > fd =3D qemu_open(mount->dirname, O_RDONLY); > if (fd =3D=3D -1) { > - error_setg_errno(errp, errno, "failed to open %s", mount= ->dirname); > - goto error; > + result->error =3D g_strdup_printf("failed to open: %s", > + strerror(errno)); > + result->has_error =3D true; > + continue; > } > = > /* We try to cull filesytems we know won't work in advance, = but other > * filesytems may not implement fstrim for less obvious reas= ons. These > - * will report EOPNOTSUPP; we simply ignore these errors. A= ny other > - * error means an unexpected error, so return it in those ca= ses. In > - * some other cases ENOTTY will be reported (e.g. CD-ROMs). > + * will report EOPNOTSUPP; while in some other cases ENOTTY = will be > + * reported (e.g. CD-ROMs). > + * Any other error means an unexpected error. > */ > r.start =3D 0; > r.len =3D -1; > r.minlen =3D has_minimum ? minimum : 0; > ret =3D ioctl(fd, FITRIM, &r); > if (ret =3D=3D -1) { > - if (errno !=3D ENOTTY && errno !=3D EOPNOTSUPP) { > - error_setg_errno(errp, errno, "failed to trim %s", > - mount->dirname); > - close(fd); > - goto error; > + result->has_error =3D true; > + if (errno =3D=3D ENOTTY || errno =3D=3D EOPNOTSUPP) { > + result->error =3D g_strdup("trim not supported"); > + } else { > + result->error =3D g_strdup_printf("failed to trim: %= s", > + strerror(errno)); > } > + close(fd); > + continue; > } > + > + result->has_minimum =3D true; > + result->minimum =3D r.minlen; > + result->has_trimmed =3D true; > + result->trimmed =3D r.len; > close(fd); > } > = > -error: > free_fs_mount_list(&mounts); > + return response; > } > #endif /* CONFIG_FSTRIM */ > = > @@ -2401,9 +2425,11 @@ int64_t qmp_guest_fsfreeze_thaw(Error **errp) > #endif /* CONFIG_FSFREEZE */ > = > #if !defined(CONFIG_FSTRIM) > -void qmp_guest_fstrim(bool has_minimum, int64_t minimum, Error **err= p) > +GuestFilesystemTrimResponse * > +qmp_guest_fstrim(bool has_minimum, int64_t minimum, Error **errp) > { > error_set(errp, QERR_UNSUPPORTED); > + return NULL; > } > #endif > = > diff --git a/qga/commands-win32.c b/qga/commands-win32.c > index 3ef0549..cc407f3 100644 > --- a/qga/commands-win32.c > +++ b/qga/commands-win32.c > @@ -493,9 +493,11 @@ static void guest_fsfreeze_cleanup(void) > * Walk list of mounted file systems in the guest, and discard unused > * areas. > */ > -void qmp_guest_fstrim(bool has_minimum, int64_t minimum, Error **err= p) > +GuestFilesystemTrimResponse * > +qmp_guest_fstrim(bool has_minimum, int64_t minimum, Error **errp) > { > error_set(errp, QERR_UNSUPPORTED); > + return NULL; > } > = > typedef enum { > diff --git a/qga/qapi-schema.json b/qga/qapi-schema.json > index 95f49e3..b4f4b93 100644 > --- a/qga/qapi-schema.json > +++ b/qga/qapi-schema.json > @@ -425,6 +425,30 @@ > 'returns': 'int' } > = > ## > +# @GuestFilesystemTrimResult > +# > +# @path: path that was trimmed > = > The path means the mount point of fs? > = > +# @error: an error message when trim failed > +# @trimmed: bytes trimmed for this path > +# @minimum: reported effective minimum for this path > = > +# > +# Since: 2.4 > +## > +{ 'type': 'GuestFilesystemTrimResult', > + 'data': {'path': 'str', > + '*trimmed': 'int', '*minimum': 'int', '*error': 'str'} } > = > Actually i am not quite sure about the whole structure, because when some= one > decides to implement > this command for Windows they may face some difficulties. > For Win 8 we have only ioctl=C2=A0 FSCTL_FILE_LEVEL_TRIM, that allows as = to specify > https://msdn.microsoft.com/en-us/library/windows/desktop/hh447300(v=3Dvs.= 85).aspx > - Offset, in bytes, from the start of the file for the range to be trimme= d. > - Length, in bytes, for the range to be trimmed. > = > And as output we have > Contains the number of ranges that were successfully processed. > If as path fs mountpoint will be used, then what should be done with other > variables? From=20what I can tell, we pass in an array of FILE_LEVEL_TRIM_RANGE, wherein the offset/length is specified. These are handled in order, so when we get back the number of ranges completed, N, we can sum up the lengths of each range up until N. Anything after that was not completed, so we can set each error to to GetLastError, or perhaps set N+1 mount's error to that, and trimmed =3D 0 for the rest (with no error, since they were simply skipped according to docs) Not sure where minimum would fit in, but simply not setting it is acceptable if I'm not too out of touch with current guidelines for schemas. So it seems like it would be workable for w32 implementations. > = > +# > = > +## > +# @GuestFilesystemTrimResponse > +# > +# @paths: list of @GuestFilesystemTrimResult per path that was trimm= ed > +# > +# Since: 2.4 > +## > +{ 'type': 'GuestFilesystemTrimResponse', > + 'data': {'paths': ['GuestFilesystemTrimResult']} } > = > Afaik instead of type, struct should be used Yes, as of: commit 895a2a80e0e054f0d5d3715aa93d10d15e49f9f7 Author: Eric Blake Date: Mon May 4 09:05:27 2015 -0600 qapi: Use 'struct' instead of 'type' in schema > = > +## > # @guest-fstrim: > # > # Discard (or "trim") blocks which are not in use by the filesystem. > @@ -437,12 +461,14 @@ > # fragmented free space, although not all blocks will be disca= rded. > # The default value is zero, meaning "discard every free block= ". > # > -# Returns: Nothing. > +# Returns: A @GuestFilesystemTrimResponse which contains the > +# status of all trimmed paths. > # > -# Since: 1.2 > +# Since: 2.4 > ## > { 'command': 'guest-fstrim', > - 'data': { '*minimum': 'int' } } > + 'data': { '*minimum': 'int' }, > + 'returns': 'GuestFilesystemTrimResponse' } > = > ## > # @guest-suspend-disk > = >=20