Skip to contentSkip to conversation

Add default config file for mkfs.xfs

RFC PATCH In review Lukas Herbolt 5 participants 11 messages May 14 – May 20

lore
Share & tools
lore
#0 lore raw

cover Add default config file for mkfs.xfs

Hi all,
The default values cannot always satisfy everybody and users/packagers should be able to override defaults without the need of specifying configuration file on command line. It was previously discussed in here [1].

The proposed patch adds /etc/mkfs.xfs.conf where the default values can be set. It also lightly changes the default behavior for -l/-d/-r concurrency allowing now to be set to -1 to keep autodetection on SSDs.

[1]
https://lore.kernel.org/linux-xfs/[email protected]/

Lukas Herbolt (1):
xfsprogs: mkfs.xfs add default configuration file.

 include/builddefs.in |  1 +
 mkfs/Makefile        |  3 ++
 mkfs/mkfs.xfs.conf   | 50 ++++++++++++++++++++++++++++++
 mkfs/xfs_mkfs.c      | 72 +++++++++++++++++++++++++++++++++++++++-----
 4 files changed, 119 insertions(+), 7 deletions(-)
 create mode 100644 mkfs/mkfs.xfs.conf
Signature · 1 line
2.54.0
#1 lore raw

RFC PATCH 1/1 xfsprogs: mkfs.xfs add default configuration file.

Various users may prefer different default values. Having a default config file will allow them to utilize it without the need specifying configuration file on command line.

---
 include/builddefs.in |  1 +
 mkfs/Makefile        |  3 ++
 mkfs/mkfs.xfs.conf   | 50 ++++++++++++++++++++++++++++++
 mkfs/xfs_mkfs.c      | 72 +++++++++++++++++++++++++++++++++++++++-----
 4 files changed, 119 insertions(+), 7 deletions(-)
 create mode 100644 mkfs/mkfs.xfs.conf
include/builddefs.in +1 -0
index 3b52d1afd703..b635a7cd08a6 100644
@@ -59,6 +59,7 @@ PKG_DOC_DIR	= @datadir@/doc/@pkg_name@
 PKG_LOCALE_DIR	= @datadir@/locale
 PKG_DATA_DIR	= @datadir@/@pkg_name@
 MKFS_CFG_DIR	= @datadir@/@pkg_name@/mkfs
+MKFS_SYSCONF_DIR = @sysconfdir@
 PKG_STATE_DIR	= @localstatedir@/lib/@pkg_name@
 
 XFS_SCRUB_ALL_AUTO_MEDIA_SCAN_STAMP=$(PKG_STATE_DIR)/xfs_scrub_all_media.stamp
mkfs/Makefile +3 -0
index fb1473324cde..57cee687eb1e 100644
@@ -21,6 +21,7 @@ CFGFILES = \
 	lts_6.12.conf \
 	lts_6.18.conf
 
+LCFLAGS += -DMKFS_CFG_DIR=\"$(MKFS_CFG_DIR)\" -DMKFS_SYSCONF_DIR=\"$(MKFS_SYSCONF_DIR)\"
 LLDLIBS += $(LIBXFS) $(LIBXCMD) $(LIBFROG) $(LIBRT) $(LIBBLKID) \
 	$(LIBUUID) $(LIBINIH) $(LIBURCU) $(LIBPTHREAD)
 LTDEPENDENCIES += $(LIBXFS) $(LIBXCMD) $(LIBFROG)
@@ -45,6 +46,8 @@ install: default
 	$(INSTALL) -m 755 $(XFS_PROTOFILE) $(PKG_SBIN_DIR)/xfs_protofile
 	$(INSTALL) -m 755 -d $(MKFS_CFG_DIR)
 	$(INSTALL) -m 644 $(CFGFILES) $(MKFS_CFG_DIR)
+	$(INSTALL) -m 755 -d $(MKFS_SYSCONF_DIR)
+	$(INSTALL) -m 644 mkfs.xfs.conf $(MKFS_SYSCONF_DIR)
 
 install-dev:
 
mkfs/mkfs.xfs.conf new file +50 -0

50 lines not on this page · raw message

mkfs/xfs_mkfs.c +65 -7

183 lines not on this page · raw message

Signature · 1 line
2.54.0
#2 lore raw

Re: [RFC PATCH 1/1] xfsprogs: mkfs.xfs add default configuration file.

Quoted 4 lines · Lukas Herbolt

Various users may prefer different default values. Having a default config file will allow them to utilize it without the need specifying configuration file on command line.

The idea seems reasonable, to have a default file to load if it's found. I just particularly don't like the idea of shipping/installing a 'default' config file.
I think we could ship an "example" one, but leave to distributions to decide what to do. If they would install a default config or not. And maintain the default config file.
LTS configs are easy to maintain because we know no new features will be ported to LTS. Providing a default config file from the mainline risks getting a lot of user complains if we in the future decide to change the 'defaults' of the default config file.

Quoted 306 lines
---
 include/builddefs.in |  1 +
 mkfs/Makefile        |  3 ++
 mkfs/mkfs.xfs.conf   | 50 ++++++++++++++++++++++++++++++
 mkfs/xfs_mkfs.c      | 72 +++++++++++++++++++++++++++++++++++++++-----
 4 files changed, 119 insertions(+), 7 deletions(-)
 create mode 100644 mkfs/mkfs.xfs.conf
include/builddefs.in +1 -0
index 3b52d1afd703..b635a7cd08a6 100644
@@ -59,6 +59,7 @@ PKG_DOC_DIR	= @datadir@/doc/@pkg_name@
 PKG_LOCALE_DIR	= @datadir@/locale
 PKG_DATA_DIR	= @datadir@/@pkg_name@
 MKFS_CFG_DIR	= @datadir@/@pkg_name@/mkfs
+MKFS_SYSCONF_DIR = @sysconfdir@
 PKG_STATE_DIR	= @localstatedir@/lib/@pkg_name@
 
 XFS_SCRUB_ALL_AUTO_MEDIA_SCAN_STAMP=$(PKG_STATE_DIR)/xfs_scrub_all_media.stamp
mkfs/Makefile +3 -0
index fb1473324cde..57cee687eb1e 100644
@@ -21,6 +21,7 @@ CFGFILES = \
 	lts_6.12.conf \
 	lts_6.18.conf
 
+LCFLAGS += -DMKFS_CFG_DIR=\"$(MKFS_CFG_DIR)\" -DMKFS_SYSCONF_DIR=\"$(MKFS_SYSCONF_DIR)\"
 LLDLIBS += $(LIBXFS) $(LIBXCMD) $(LIBFROG) $(LIBRT) $(LIBBLKID) \
 	$(LIBUUID) $(LIBINIH) $(LIBURCU) $(LIBPTHREAD)
 LTDEPENDENCIES += $(LIBXFS) $(LIBXCMD) $(LIBFROG)
@@ -45,6 +46,8 @@ install: default
 	$(INSTALL) -m 755 $(XFS_PROTOFILE) $(PKG_SBIN_DIR)/xfs_protofile
 	$(INSTALL) -m 755 -d $(MKFS_CFG_DIR)
 	$(INSTALL) -m 644 $(CFGFILES) $(MKFS_CFG_DIR)
+	$(INSTALL) -m 755 -d $(MKFS_SYSCONF_DIR)
+	$(INSTALL) -m 644 mkfs.xfs.conf $(MKFS_SYSCONF_DIR)
 
 install-dev:
 
mkfs/mkfs.xfs.conf new file +50 -0
new file mode 100644
index 000000000000..76f5ab4d4d8e
@@ -0,0 +1,50 @@
+# mkfs.xfs default configuration file
+#
+# This file documents some of the options recognised by mkfs.xfs config file.
+# Adjust any value to override it system-wide.
+#
+# Command-line options always take precedence over values in this file.
+# A specific config file can be selected with: mkfs.xfs -c options=<path>
+
+[block]
+#size = 4096
+
+[metadata]
+#crc = 1
+#finobt = 1
+#inobtcount = 1
+#rmapbt = 1
+#reflink = 1
+#bigtime = 1
+#metadir = 0
+#autofsck = 0
+
+[inode]
+#align = 1
+# The default value is 25% for filesystems under 1TB,#5% for filesystems under
+# 50TB and 1% for filesystems over 50TB.
+#
+#maxpct = 25
+#size = 512
+#perblock = 8
+#attr = 2
+#projid32bit = 1
+#sparse = 1
+#nrext64 = 1
+#exchange = 1
+
+[data]
+# -1 = autodetect (use CPU count only on solid-state devices), 0 = disabled
+#concurrency = -1
+
+[log]
+#internal = 1
+#version = 2
+#lazy-count = 1
+# -1 = autodetect (use CPU count only on solid-state devices), 0 = disabled
+#concurrency = -1
+
+[realtime]
+# -1 = autodetect (use CPU count only on solid-state devices), 0 = disabled
+#concurrency = -1
+
mkfs/xfs_mkfs.c +65 -7
index dd8a48c3633e..dbf15eca3442 100644
@@ -44,6 +44,11 @@
  */
 #define WHACK_SIZE (128 * 1024)
 
+/*
+ * Default configuration file which can keep distro specific values.
+ */
+#define MKFS_DEFAULT_CFGFILE	MKFS_SYSCONF_DIR "/mkfs.xfs.conf"
+
 /*
  * XXX: The configured block and sector sizes are defined as global variables so
  * that they don't need to be passed to getnum/cvtnum().
@@ -51,6 +56,11 @@
 static unsigned int		blocksize;
 static unsigned int		sectorsize;
 
+/*
+ * Set to true while parsing the config file so option handlers know the source
+ */
+static bool			parsing_cfgfile;
+
 /*
  * Enums for each CLI parameter type are declared first so we can calculate the
  * maximum array size needed to hold them automatically.
@@ -264,6 +274,7 @@ struct opt_params {
 		bool		str_seen;
 		bool		convert;
 		bool		is_power_2;
+		bool		from_file;
 		struct _conflict {
 			struct opt_params	*opts;
 			int			subopt;
@@ -472,7 +483,7 @@ static struct opt_params dopts = {
 		  .conflicts = { { &dopts, D_AGCOUNT },
 				 { &dopts, D_AGSIZE },
 				 { NULL, LAST_CONFLICT } },
-		  .minval = 0,
+		  .minval = -1,
 		  .maxval = INT_MAX,
 		  .defaultval = 1,
 		},
@@ -672,7 +683,7 @@ static struct opt_params lopts = {
 				 { &lopts, L_FILE },
 				 { &lopts, L_DEV },
 				 { NULL, LAST_CONFLICT } },
-		  .minval = 0,
+		  .minval = -1,
 		  .maxval = INT_MAX,
 		  .defaultval = 1,
 		},
@@ -827,7 +838,7 @@ static struct opt_params ropts = {
 				 { &ropts, R_RGSIZE },
 				 { NULL, LAST_CONFLICT } },
 		  .convert = true,
-		  .minval = 0,
+		  .minval = -1,
 		  .maxval = INT_MAX,
 		  .defaultval = 1,
 		},
@@ -1072,6 +1083,7 @@ struct cli_params {
 	int	blocksize;
 
 	char	*cfgfile;
+	bool	cfgfile_had_options;
 	char	*protofile;
 
 	enum fsprop_autofsck autofsck;
@@ -1654,10 +1666,12 @@ check_opt(
 		if (sp->seen)
 			respec(opts->name, opts->subopts, index);
 		sp->seen = true;
+		sp->from_file = parsing_cfgfile;
 	} else {
 		if (sp->str_seen)
 			respec(opts->name, opts->subopts, index);
 		sp->str_seen = true;
+		sp->from_file = parsing_cfgfile;
 	}
 
 	/* check for conflicts with the option */
@@ -1806,6 +1820,7 @@ set_data_concurrency(
 	/*
 	 * "nr_cpus" or "1" means set the concurrency level to the CPU count.
 	 * If this cannot be determined, fall back to the default AG geometry.
+	 * -1 means autodetect (use CPU count only on solid-state devices).
 	 */
 	if (!value || !strcmp(value, "nr_cpus"))
 		optnum = 1;
@@ -1814,6 +1829,8 @@ set_data_concurrency(
 
 	if (optnum == 1)
 		cli->data_concurrency = nr_cpus();
+	else if (optnum == -1)
+		cli->data_concurrency = -1;
 	else
 		cli->data_concurrency = optnum;
 }
@@ -1954,6 +1971,7 @@ set_log_concurrency(
 	/*
 	 * "nr_cpus" or 1 means set the concurrency level to the CPU count.  If
 	 * this cannot be determined, fall back to the default computation.
+	 * -1 means autodetect (use CPU count only on solid-state devices).
 	 */
 	if (!value || !strcmp(value, "nr_cpus"))
 		optnum = 1;
@@ -1962,6 +1980,8 @@ set_log_concurrency(
 
 	if (optnum == 1)
 		cli->log_concurrency = nr_cpus();
+	else if (optnum == -1)
+		cli->log_concurrency = -1;
 	else
 		cli->log_concurrency = optnum;
 }
@@ -2175,6 +2195,7 @@ set_rtvol_concurrency(
 	 * "nr_cpus" or "1" means set the concurrency level to the CPU count.
 	 * If this cannot be determined, fall back to the default rtgroup
 	 * geometry.
+	 * -1 means autodetect (use CPU count only on solid-state devices).
 	 */
 	if (!value || !strcmp(value, "nr_cpus"))
 		optnum = 1;
@@ -2183,6 +2204,8 @@ set_rtvol_concurrency(
 
 	if (optnum == 1)
 		cli->rtvol_concurrency = nr_cpus();
+	else if (optnum == -1)
+		cli->rtvol_concurrency = -1;
 	else
 		cli->rtvol_concurrency = optnum;
 }
@@ -2336,9 +2359,32 @@ parse_cfgopt(
 		if (!subopts[i])
 			break;
 		if (strcasecmp(name, subopts[i]) == 0) {
+			struct subopt_param	*sp = &sop->opts->subopt_params[i];
+			int			j;
+
+			/*
+			 * Command line options take precedence over config file
+			 * options. If this option or any option that conflicts
+			 * with it was already set from the command line, skip
+			 * the config file value silently.
+			 */
+			if ((sp->seen || sp->str_seen) && !sp->from_file)
+				return true;
+			for (j = 0; j < MAX_CONFLICTS; j++) {
+				struct _conflict	*con = &sp->conflicts[j];
+				struct subopt_param	*csp;
+
+				if (con->subopt == LAST_CONFLICT)
+					break;
+				csp = &con->opts->subopt_params[con->subopt];
+				if ((csp->seen || csp->str_seen) && !csp->from_file)
+					return true;
+			}
+
 			ret = (sop->parser)(sop->opts, i, value, cli);
 			if (ret)
 				goto invalid_opt;
+			cli->cfgfile_had_options = true;
 			return true;
 		}
 	}
@@ -5749,10 +5795,21 @@ cfgfile_parse(
 {
 	int			error;
 
-	if (!cli->cfgfile)
-		return;
+	bool	default_cfgfile = !cli->cfgfile;
+
+	if (!cli->cfgfile) {
+		/*
+		 * No config file specified on the command line. Use the default
+		 * system-wide config file if it exists, otherwise do nothing.
+		 */
+		if (access(MKFS_DEFAULT_CFGFILE, F_OK) != 0)
+			return;
+		cli->cfgfile = MKFS_DEFAULT_CFGFILE;
+	}
 
+	parsing_cfgfile = true;
 	error = ini_parse(cli->cfgfile, cfgfile_parse_ini, cli);
+	parsing_cfgfile = false;
 	if (error) {
 		if (error > 0) {
 			fprintf(stderr,
@@ -5773,8 +5830,9 @@ cfgfile_parse(
 		}
 		exit(1);
 	}
-	printf(_("Parameters parsed from config file %s successfully\n"),
-		cli->cfgfile);
+	if (!default_cfgfile || cli->cfgfile_had_options)
+		printf(_("Parameters parsed from config file %s successfully\n"),
+			cli->cfgfile);
 }
 
 static void
Signature · 1 line
2.54.0
#3 lore raw

Re: [RFC PATCH 1/1] xfsprogs: mkfs.xfs add default configuration file.

Quoted 5 lines · Carlos Maiolino
Quoted 4 lines · Lukas Herbolt

Various users may prefer different default values. Having a default config file will allow them to utilize it without the need specifying configuration file on command line.

Quoted 10 lines

The idea seems reasonable, to have a default file to load if it's found. I just particularly don't like the idea of shipping/installing a 'default' config file.
I think we could ship an "example" one, but leave to distributions to decide what to do. If they would install a default config or not. And maintain the default config file.
LTS configs are easy to maintain because we know no new features will be ported to LTS. Providing a default config file from the mainline risks getting a lot of user complains if we in the future decide to change the 'defaults' of the default config file.

Thanks for starting this discussion Lukas.

I'm sure others will have Deeper Thoughts but I more or less agree with Carlos here - a default config file to be loaded /if it exists/ but maybe not existing by default - because we already have defaults hardcoded into the binary itself - might be a more maintainable path for upstream. Otherwise I see new tests appearing which are like "make sure the dfault config we ship matches the defaults built into the binary itself" which seems like unneeded complexity.

(i.e. the os-specific config files could be renamed or symlinked to that special default config file path, for example?)

I could maybe see generating an example defaults config file which matches the built-in defaults at build time? Not sure if that would be useful or helpful (or complex).

Thanks,
-Eric

#4 lore raw

Re: [RFC PATCH 1/1] xfsprogs: mkfs.xfs add default configuration file.

Quoted 27 lines · Eric Sandeen
Quoted 5 lines · Carlos Maiolino
Quoted 4 lines · Lukas Herbolt

Various users may prefer different default values. Having a default config file will allow them to utilize it without the need specifying configuration file on command line.

Quoted 10 lines

The idea seems reasonable, to have a default file to load if it's found. I just particularly don't like the idea of shipping/installing a 'default' config file.
I think we could ship an "example" one, but leave to distributions to decide what to do. If they would install a default config or not. And maintain the default config file.
LTS configs are easy to maintain because we know no new features will be ported to LTS. Providing a default config file from the mainline risks getting a lot of user complains if we in the future decide to change the 'defaults' of the default config file.

Thanks for starting this discussion Lukas.

I'm sure others will have Deeper Thoughts but I more or less agree with Carlos here - a default config file to be loaded /if it exists/ but maybe not existing by default - because we already have defaults hardcoded into the binary itself - might be a more maintainable path for upstream. Otherwise I see new tests appearing which are like "make sure the dfault config we ship matches the defaults built into the binary itself" which seems like unneeded complexity.

Quoted 6 lines

(i.e. the os-specific config files could be renamed or symlinked to that special default config file path, for example?)

I could maybe see generating an example defaults config file which matches the built-in defaults at build time? Not sure if that would be useful or helpful (or complex).

I think that would be a lot easier than us trying to keep the default config file in sync with xfs_mkfs.c by hand.

--D

Quoted 4 lines

Thanks,
-Eric

#5 lore raw

Re: [RFC PATCH 1/1] xfsprogs: mkfs.xfs add default configuration file.

In reply to Eric Sandeen
Quoted 17 lines · Eric Sandeen
Quoted 5 lines · Carlos Maiolino
Quoted 4 lines · Lukas Herbolt

Various users may prefer different default values. Having a default config file will allow them to utilize it without the need specifying configuration file on command line.

Quoted 10 lines

The idea seems reasonable, to have a default file to load if it's found. I just particularly don't like the idea of shipping/installing a 'default' config file.
I think we could ship an "example" one, but leave to distributions to decide what to do. If they would install a default config or not. And maintain the default config file.
LTS configs are easy to maintain because we know no new features will be ported to LTS. Providing a default config file from the mainline risks getting a lot of user complains if we in the future decide to change the 'defaults' of the default config file.

Quoted 6 lines

Thanks for starting this discussion Lukas.

I'm sure others will have Deeper Thoughts but I more or less agree with Carlos here - a default config file to be loaded /if it exists/ but maybe not existing by default - because we already have defaults hardcoded into the binary itself - might be a more maintainable path for upstream.

Yes, the upstream package should not install a defaults file - the package might be built alongside the system package install using an install prefix, and we don't want to be overriding any config the user/distro has already installed.

Quoted 10 lines

Otherwise I see new tests appearing which are like "make sure the dfault config we ship matches the defaults built into the binary itself" which seems like unneeded complexity.

(i.e. the os-specific config files could be renamed or symlinked to that special default config file path, for example?)

I could maybe see generating an example defaults config file which matches the built-in defaults at build time? Not sure if that would be useful or helpful (or complex).

Not really useful, IMO, because changing default behaviour is a relatively rare event.

Indeed, we already have LTS kernel mkfs conf files that adjust "cli" parameters to match the mkfs defaults at the time the LTS kernel was released. These get installed into /usr/share/xfsprogs/mkfs for users that want to use them via 'mkfs.xfs -c ..../lts_6.12.conf ...'.

Hence I don't think we even need to care about maintaining conf files for the current defaults or even per-package-release files because what most people care about is maintaining a feature set that is compatible with the distro kernel(s) they are running...

Hence if we are going to ship default config files that users can install manually into /etc/xfs/mkfs/default.conf, then the package install directory for those example config files would also be /usr/share/xfsprogs/mkfs, right?

If that is the case, how does the user now tell the difference between a CLI conf file and a template default conf file that should be installed into /etc/xfsprogs/mkfs/.... to change the default behaviour to match lts_6.12.conf?

I guess what I'm asking is that the conf file we install into /etc to set the defaults needs to have exactly the same contents as the conf file the user would specify on the command line to get that specific config to be used.

If this is already the case with this new proposal, then we don't need to ship any new example config files, just hook up an xfs_admin command to set a specific conf file as the mkfs default. i.e. to set the default behaviour to 6.12 compatible:

# xfs_admin -d /usr/share/xfsprogs/mkfs/lts_6.12.conf

And to reset mkfs behaviour back to built in defaults:

# xfs_admin -d clear

Distros can then patch their package source to add their own defaults files for their own releases - we could even ask distros to push them upstream so that users can still use bleeding edge xfsprogs on older distros when needed....

Cheers,

Dave.

Signature · 2 lines
Dave Chinner
[email protected]
#6 lore raw

Re: [RFC PATCH 1/1] xfsprogs: mkfs.xfs add default configuration file.

In reply to Eric Sandeen
Quoted 30 lines · Eric Sandeen
Quoted 5 lines · Carlos Maiolino
Quoted 4 lines · Lukas Herbolt

Various users may prefer different default values. Having a default config file will allow them to utilize it without the need specifying configuration file on command line.

Quoted 10 lines

The idea seems reasonable, to have a default file to load if it's found. I just particularly don't like the idea of shipping/installing a 'default' config file.
I think we could ship an "example" one, but leave to distributions to decide what to do. If they would install a default config or not. And maintain the default config file.
LTS configs are easy to maintain because we know no new features will be ported to LTS. Providing a default config file from the mainline risks getting a lot of user complains if we in the future decide to change the 'defaults' of the default config file.

Thanks for starting this discussion Lukas.

I'm sure others will have Deeper Thoughts but I more or less agree with Carlos here - a default config file to be loaded /if it exists/ but maybe not existing by default - because we already have defaults hardcoded into the binary itself - might be a more maintainable path for upstream. Otherwise I see new tests appearing which are like "make sure the dfault config we ship matches the defaults built into the binary itself" which seems like unneeded complexity.

(i.e. the os-specific config files could be renamed or symlinked to that special default config file path, for example?)

Quoted 6 lines

I could maybe see generating an example defaults config file which matches the built-in defaults at build time? Not sure if that would be useful or helpful (or complex).

Thanks,
-Eric

Noted, I think we can be shipping mkfs.xfs.conf.example with some of the defaults options set and keep it with the lts configs

Signature · 1 line
-lhe
#7 lore raw

Re: [RFC PATCH 1/1] xfsprogs: mkfs.xfs add default configuration file.

In reply to Lukas Herbolt
Quoted 38 lines · Lukas Herbolt

Various users may prefer different default values. Having a default config file will allow them to utilize it without the need specifying configuration file on command line.

---
 include/builddefs.in |  1 +
 mkfs/Makefile        |  3 ++
 mkfs/mkfs.xfs.conf   | 50 ++++++++++++++++++++++++++++++
 mkfs/xfs_mkfs.c      | 72 +++++++++++++++++++++++++++++++++++++++-----
 4 files changed, 119 insertions(+), 7 deletions(-)
 create mode 100644 mkfs/mkfs.xfs.conf
include/builddefs.in +1 -0
index 3b52d1afd703..b635a7cd08a6 100644
@@ -59,6 +59,7 @@ PKG_DOC_DIR	= @datadir@/doc/@pkg_name@
 PKG_LOCALE_DIR	= @datadir@/locale
 PKG_DATA_DIR	= @datadir@/@pkg_name@
 MKFS_CFG_DIR	= @datadir@/@pkg_name@/mkfs
+MKFS_SYSCONF_DIR = @sysconfdir@
 PKG_STATE_DIR	= @localstatedir@/lib/@pkg_name@
 
 XFS_SCRUB_ALL_AUTO_MEDIA_SCAN_STAMP=$(PKG_STATE_DIR)/xfs_scrub_all_media.stamp
mkfs/Makefile +1 -0
index fb1473324cde..57cee687eb1e 100644
@@ -21,6 +21,7 @@ CFGFILES = \
 	lts_6.12.conf \
 	lts_6.18.conf
 
+LCFLAGS += -DMKFS_CFG_DIR=\"$(MKFS_CFG_DIR)\" -DMKFS_SYSCONF_DIR=\"$(MKFS_SYSCONF_DIR)\"
 LLDLIBS += $(LIBXFS) $(LIBXCMD) $(LIBFROG) $(LIBRT) $(LIBBLKID) \
 	$(LIBUUID) $(LIBINIH) $(LIBURCU) $(LIBPTHREAD)
 LTDEPENDENCIES += $(LIBXFS) $(LIBXCMD) $(LIBFROG)

mkfs/Makefile install: default · line 50

Quoted 5 lines
mkfs/Makefile +2 -0
@@ -45,6 +46,8 @@ install: default
 	$(INSTALL) -m 755 $(XFS_PROTOFILE) $(PKG_SBIN_DIR)/xfs_protofile
 	$(INSTALL) -m 755 -d $(MKFS_CFG_DIR)
 	$(INSTALL) -m 644 $(CFGFILES) $(MKFS_CFG_DIR)
+	$(INSTALL) -m 755 -d $(MKFS_SYSCONF_DIR)
+	$(INSTALL) -m 644 mkfs.xfs.conf $(MKFS_SYSCONF_DIR)

At first I thought "gee, are we not supposed to dump config files into /etc anymore?" but maybe that's a crazy ro-root/systemd thing. I guess that at least it /is/ in the spirit of /etc/mke2fs.conf.

Quoted 47 lines
mkfs/Makefile +0 -0
 install-dev:
 
mkfs/mkfs.xfs.conf new file +39 -0
new file mode 100644
index 000000000000..76f5ab4d4d8e
@@ -0,0 +1,50 @@
+# mkfs.xfs default configuration file
+#
+# This file documents some of the options recognised by mkfs.xfs config file.
+# Adjust any value to override it system-wide.
+#
+# Command-line options always take precedence over values in this file.
+# A specific config file can be selected with: mkfs.xfs -c options=<path>
+
+[block]
+#size = 4096
+
+[metadata]
+#crc = 1
+#finobt = 1
+#inobtcount = 1
+#rmapbt = 1
+#reflink = 1
+#bigtime = 1
+#metadir = 0
+#autofsck = 0
+
+[inode]
+#align = 1
+# The default value is 25% for filesystems under 1TB,#5% for filesystems under
+# 50TB and 1% for filesystems over 50TB.
+#
+#maxpct = 25
+#size = 512
+#perblock = 8
+#attr = 2
+#projid32bit = 1
+#sparse = 1
+#nrext64 = 1
+#exchange = 1
+
+[data]
+# -1 = autodetect (use CPU count only on solid-state devices), 0 = disabled
+#concurrency = -1
+

mkfs/mkfs.xfs.conf line 49

Quoted 10 lines
mkfs/mkfs.xfs.conf new file +10 -0
+[log]
+#internal = 1
+#version = 2
+#lazy-count = 1
+# -1 = autodetect (use CPU count only on solid-state devices), 0 = disabled
+#concurrency = -1
+
+[realtime]
+# -1 = autodetect (use CPU count only on solid-state devices), 0 = disabled
+#concurrency = -1

Oh, I see. The "-1" changes scattered throughout are to make it so that you can specify that as a concurrency= option. I had mistakenly thought pre-coffee that they were adding a new value.

However, I think "concurrency=auto" would be more ergonomic for users.

	/*
	 * "nr_cpus" or "1" means set the concurrency level to the CPU count.
	 * If this cannot be determined, fall back to the default AG geometry.
	 * "auto" means use CPU count only solid-state devices
	 */
	if (!value || !strcmp(value, "nr_cpus"))
		optnum = 1;
	else if (!strcmp(value, "auto"))
		optnum = -1;
	else
		optnum = getnum(value, opts, subopt);

	if (optnum == 1)
		cli->data_concurrency = nr_cpus();
	else
		cli->data_concurrency = optnum;

Also note that the explicit strcmp carveout for "auto" (and "nr_cpus") means you don't have to tweak minval below.

Whatever we end up adding for a "just use the defaults" value, that ought to be a separate patch from the one that adds the defaults config file.

Quoted 137 lines
mkfs/xfs_mkfs.c +26 -3
index dd8a48c3633e..dbf15eca3442 100644
@@ -44,6 +44,11 @@
  */
 #define WHACK_SIZE (128 * 1024)
 
+/*
+ * Default configuration file which can keep distro specific values.
+ */
+#define MKFS_DEFAULT_CFGFILE	MKFS_SYSCONF_DIR "/mkfs.xfs.conf"
+
 /*
  * XXX: The configured block and sector sizes are defined as global variables so
  * that they don't need to be passed to getnum/cvtnum().
@@ -51,6 +56,11 @@
 static unsigned int		blocksize;
 static unsigned int		sectorsize;
 
+/*
+ * Set to true while parsing the config file so option handlers know the source
+ */
+static bool			parsing_cfgfile;
+
 /*
  * Enums for each CLI parameter type are declared first so we can calculate the
  * maximum array size needed to hold them automatically.
@@ -264,6 +274,7 @@ struct opt_params {
 		bool		str_seen;
 		bool		convert;
 		bool		is_power_2;
+		bool		from_file;
 		struct _conflict {
 			struct opt_params	*opts;
 			int			subopt;
@@ -472,7 +483,7 @@ static struct opt_params dopts = {
 		  .conflicts = { { &dopts, D_AGCOUNT },
 				 { &dopts, D_AGSIZE },
 				 { NULL, LAST_CONFLICT } },
-		  .minval = 0,
+		  .minval = -1,
 		  .maxval = INT_MAX,
 		  .defaultval = 1,
 		},
@@ -672,7 +683,7 @@ static struct opt_params lopts = {
 				 { &lopts, L_FILE },
 				 { &lopts, L_DEV },
 				 { NULL, LAST_CONFLICT } },
-		  .minval = 0,
+		  .minval = -1,
 		  .maxval = INT_MAX,
 		  .defaultval = 1,
 		},
@@ -827,7 +838,7 @@ static struct opt_params ropts = {
 				 { &ropts, R_RGSIZE },
 				 { NULL, LAST_CONFLICT } },
 		  .convert = true,
-		  .minval = 0,
+		  .minval = -1,
 		  .maxval = INT_MAX,
 		  .defaultval = 1,
 		},
@@ -1072,6 +1083,7 @@ struct cli_params {
 	int	blocksize;
 
 	char	*cfgfile;
+	bool	cfgfile_had_options;
 	char	*protofile;
 
 	enum fsprop_autofsck autofsck;
@@ -1654,10 +1666,12 @@ check_opt(
 		if (sp->seen)
 			respec(opts->name, opts->subopts, index);
 		sp->seen = true;
+		sp->from_file = parsing_cfgfile;
 	} else {
 		if (sp->str_seen)
 			respec(opts->name, opts->subopts, index);
 		sp->str_seen = true;
+		sp->from_file = parsing_cfgfile;
 	}
 
 	/* check for conflicts with the option */
@@ -1806,6 +1820,7 @@ set_data_concurrency(
 	/*
 	 * "nr_cpus" or "1" means set the concurrency level to the CPU count.
 	 * If this cannot be determined, fall back to the default AG geometry.
+	 * -1 means autodetect (use CPU count only on solid-state devices).
 	 */
 	if (!value || !strcmp(value, "nr_cpus"))
 		optnum = 1;
@@ -1814,6 +1829,8 @@ set_data_concurrency(
 
 	if (optnum == 1)
 		cli->data_concurrency = nr_cpus();
+	else if (optnum == -1)
+		cli->data_concurrency = -1;
 	else
 		cli->data_concurrency = optnum;
 }
@@ -1954,6 +1971,7 @@ set_log_concurrency(
 	/*
 	 * "nr_cpus" or 1 means set the concurrency level to the CPU count.  If
 	 * this cannot be determined, fall back to the default computation.
+	 * -1 means autodetect (use CPU count only on solid-state devices).
 	 */
 	if (!value || !strcmp(value, "nr_cpus"))
 		optnum = 1;
@@ -1962,6 +1980,8 @@ set_log_concurrency(
 
 	if (optnum == 1)
 		cli->log_concurrency = nr_cpus();
+	else if (optnum == -1)
+		cli->log_concurrency = -1;
 	else
 		cli->log_concurrency = optnum;
 }
@@ -2175,6 +2195,7 @@ set_rtvol_concurrency(
 	 * "nr_cpus" or "1" means set the concurrency level to the CPU count.
 	 * If this cannot be determined, fall back to the default rtgroup
 	 * geometry.
+	 * -1 means autodetect (use CPU count only on solid-state devices).
 	 */
 	if (!value || !strcmp(value, "nr_cpus"))
 		optnum = 1;
@@ -2183,6 +2204,8 @@ set_rtvol_concurrency(
 
 	if (optnum == 1)
 		cli->rtvol_concurrency = nr_cpus();
+	else if (optnum == -1)
+		cli->rtvol_concurrency = -1;
 	else
 		cli->rtvol_concurrency = optnum;
 }
@@ -2336,9 +2359,32 @@ parse_cfgopt(
 		if (!subopts[i])

mkfs/xfs_mkfs.c parse_cfgopt( · line 2369

Quoted 10 lines
mkfs/xfs_mkfs.c +8 -0
 			break;
 		if (strcasecmp(name, subopts[i]) == 0) {
+			struct subopt_param	*sp = &sop->opts->subopt_params[i];
+			int			j;
+
+			/*
+			 * Command line options take precedence over config file
+			 * options. If this option or any option that conflicts
+			 * with it was already set from the command line, skip
+			 * the config file value silently.

Currently, config file options take precedence over CLI options:

$ mkfs.xfs -c options=/usr/share/xfsprogs/mkfs/lts_6.6.conf -i nrext64=0 /tmp/a
-i nrext64 option respecified

So I guess this patch is changing that policy too?

--D

Quoted 61 lines
mkfs/xfs_mkfs.c +31 -4
+			 */
+			if ((sp->seen || sp->str_seen) && !sp->from_file)
+				return true;
+			for (j = 0; j < MAX_CONFLICTS; j++) {
+				struct _conflict	*con = &sp->conflicts[j];
+				struct subopt_param	*csp;
+
+				if (con->subopt == LAST_CONFLICT)
+					break;
+				csp = &con->opts->subopt_params[con->subopt];
+				if ((csp->seen || csp->str_seen) && !csp->from_file)
+					return true;
+			}
+
 			ret = (sop->parser)(sop->opts, i, value, cli);
 			if (ret)
 				goto invalid_opt;
+			cli->cfgfile_had_options = true;
 			return true;
 		}
 	}
@@ -5749,10 +5795,21 @@ cfgfile_parse(
 {
 	int			error;
 
-	if (!cli->cfgfile)
-		return;
+	bool	default_cfgfile = !cli->cfgfile;
+
+	if (!cli->cfgfile) {
+		/*
+		 * No config file specified on the command line. Use the default
+		 * system-wide config file if it exists, otherwise do nothing.
+		 */
+		if (access(MKFS_DEFAULT_CFGFILE, F_OK) != 0)
+			return;
+		cli->cfgfile = MKFS_DEFAULT_CFGFILE;
+	}
 
+	parsing_cfgfile = true;
 	error = ini_parse(cli->cfgfile, cfgfile_parse_ini, cli);
+	parsing_cfgfile = false;
 	if (error) {
 		if (error > 0) {
 			fprintf(stderr,
@@ -5773,8 +5830,9 @@ cfgfile_parse(
 		}
 		exit(1);
 	}
-	printf(_("Parameters parsed from config file %s successfully\n"),
-		cli->cfgfile);
+	if (!default_cfgfile || cli->cfgfile_had_options)
+		printf(_("Parameters parsed from config file %s successfully\n"),
+			cli->cfgfile);
 }
 
 static void
Signature · 1 line
2.54.0
Reviewed-by Eric Sandeen on v1 1/1
#8 lore raw

Re: [RFC PATCH 1/1] xfsprogs: mkfs.xfs add default configuration file.

In reply to Lukas Herbolt
Quoted 52 lines · Lukas Herbolt

Various users may prefer different default values. Having a default config file will allow them to utilize it without the need specifying configuration file on command line.

---
 include/builddefs.in |  1 +
 mkfs/Makefile        |  3 ++
 mkfs/mkfs.xfs.conf   | 50 ++++++++++++++++++++++++++++++
 mkfs/xfs_mkfs.c      | 72 +++++++++++++++++++++++++++++++++++++++-----
 4 files changed, 119 insertions(+), 7 deletions(-)
 create mode 100644 mkfs/mkfs.xfs.conf
include/builddefs.in +1 -0
index 3b52d1afd703..b635a7cd08a6 100644
@@ -59,6 +59,7 @@ PKG_DOC_DIR	= @datadir@/doc/@pkg_name@
 PKG_LOCALE_DIR	= @datadir@/locale
 PKG_DATA_DIR	= @datadir@/@pkg_name@
 MKFS_CFG_DIR	= @datadir@/@pkg_name@/mkfs
+MKFS_SYSCONF_DIR = @sysconfdir@
 PKG_STATE_DIR	= @localstatedir@/lib/@pkg_name@
 
 XFS_SCRUB_ALL_AUTO_MEDIA_SCAN_STAMP=$(PKG_STATE_DIR)/xfs_scrub_all_media.stamp
mkfs/Makefile +3 -0
index fb1473324cde..57cee687eb1e 100644
@@ -21,6 +21,7 @@ CFGFILES = \
 	lts_6.12.conf \
 	lts_6.18.conf
 
+LCFLAGS += -DMKFS_CFG_DIR=\"$(MKFS_CFG_DIR)\" -DMKFS_SYSCONF_DIR=\"$(MKFS_SYSCONF_DIR)\"
 LLDLIBS += $(LIBXFS) $(LIBXCMD) $(LIBFROG) $(LIBRT) $(LIBBLKID) \
 	$(LIBUUID) $(LIBINIH) $(LIBURCU) $(LIBPTHREAD)
 LTDEPENDENCIES += $(LIBXFS) $(LIBXCMD) $(LIBFROG)
@@ -45,6 +46,8 @@ install: default
 	$(INSTALL) -m 755 $(XFS_PROTOFILE) $(PKG_SBIN_DIR)/xfs_protofile
 	$(INSTALL) -m 755 -d $(MKFS_CFG_DIR)
 	$(INSTALL) -m 644 $(CFGFILES) $(MKFS_CFG_DIR)
+	$(INSTALL) -m 755 -d $(MKFS_SYSCONF_DIR)
+	$(INSTALL) -m 644 mkfs.xfs.conf $(MKFS_SYSCONF_DIR)
 
 install-dev:
 
mkfs/mkfs.xfs.conf new file +0 -0
new file mode 100644
index 000000000000..76f5ab4d4d8e

mkfs/mkfs.xfs.conf line 10

Quoted 10 lines
mkfs/mkfs.xfs.conf new file +10 -0
@@ -0,0 +1,50 @@
+# mkfs.xfs default configuration file
+#
+# This file documents some of the options recognised by mkfs.xfs config file.
+# Adjust any value to override it system-wide.
+#
+# Command-line options always take precedence over values in this file.
+# A specific config file can be selected with: mkfs.xfs -c options=<path>
+
+[block]
+#size = 4096

Why do we need a config file that contains all the current defaults commented out? We don't do this for the command line conf files we ship, so this just seems like unnecessary maintenance overhead to me...

Quoted 18 lines
mkfs/mkfs.xfs.conf new file +18 -0
+
+[metadata]
+#crc = 1
+#finobt = 1
+#inobtcount = 1
+#rmapbt = 1
+#reflink = 1
+#bigtime = 1
+#metadir = 0
+#autofsck = 0
+
+[inode]
+#align = 1
+# The default value is 25% for filesystems under 1TB,#5% for filesystems under
+# 50TB and 1% for filesystems over 50TB.
+#
+#maxpct = 25
+#size = 512

mkfs/mkfs.xfs.conf line 38

Quoted 10 lines
mkfs/mkfs.xfs.conf new file +10 -0
+#perblock = 8
+#attr = 2
+#projid32bit = 1
+#sparse = 1
+#nrext64 = 1
+#exchange = 1
+
+[data]
+# -1 = autodetect (use CPU count only on solid-state devices), 0 = disabled
+#concurrency = -1

Not sure I like this way saying "autotune". This should match the existing conf file behaviour to override the default. i.e. 0 turns it off, 1 = autodetect, "nr_cpus" = autodetect", any other positive value is the minimum concurrency value.

IMO we should be exactly consistent with the CLI options here so that we don't need special parsing code just for the default options.

Quoted 16 lines
mkfs/mkfs.xfs.conf new file +11 -0
+[log]
+#internal = 1
+#version = 2
+#lazy-count = 1
+# -1 = autodetect (use CPU count only on solid-state devices), 0 = disabled
+#concurrency = -1
+
+[realtime]
+# -1 = autodetect (use CPU count only on solid-state devices), 0 = disabled
+#concurrency = -1
+
mkfs/xfs_mkfs.c +0 -0
index dd8a48c3633e..dbf15eca3442 100644

mkfs/xfs_mkfs.c line 50

Quoted 7 lines
mkfs/xfs_mkfs.c +4 -0
@@ -44,6 +44,11 @@
  */
 #define WHACK_SIZE (128 * 1024)
 
+/*
+ * Default configuration file which can keep distro specific values.
+ */
+#define MKFS_DEFAULT_CFGFILE	MKFS_SYSCONF_DIR "/mkfs.xfs.conf"

Should we make the path for MKFS_SYSCONF_DIR mkfs specific, allowing more than just one potential config file to be installed here? e.g. /etc/xfsprogs/mkfs/ allows us to use this directory for more than just a single default config file. While we probably don't need that right now, starting from a directory based conf fiel setup means we aren't stuck with having to support a single hardcoded/fixed conf file name forever.

Quoted 163 lines
mkfs/xfs_mkfs.c +46 -5
+
 /*
  * XXX: The configured block and sector sizes are defined as global variables so
  * that they don't need to be passed to getnum/cvtnum().
@@ -51,6 +56,11 @@
 static unsigned int		blocksize;
 static unsigned int		sectorsize;
 
+/*
+ * Set to true while parsing the config file so option handlers know the source
+ */
+static bool			parsing_cfgfile;
+
 /*
  * Enums for each CLI parameter type are declared first so we can calculate the
  * maximum array size needed to hold them automatically.
@@ -264,6 +274,7 @@ struct opt_params {
 		bool		str_seen;
 		bool		convert;
 		bool		is_power_2;
+		bool		from_file;
 		struct _conflict {
 			struct opt_params	*opts;
 			int			subopt;
@@ -472,7 +483,7 @@ static struct opt_params dopts = {
 		  .conflicts = { { &dopts, D_AGCOUNT },
 				 { &dopts, D_AGSIZE },
 				 { NULL, LAST_CONFLICT } },
-		  .minval = 0,
+		  .minval = -1,
 		  .maxval = INT_MAX,
 		  .defaultval = 1,
 		},
@@ -672,7 +683,7 @@ static struct opt_params lopts = {
 				 { &lopts, L_FILE },
 				 { &lopts, L_DEV },
 				 { NULL, LAST_CONFLICT } },
-		  .minval = 0,
+		  .minval = -1,
 		  .maxval = INT_MAX,
 		  .defaultval = 1,
 		},
@@ -827,7 +838,7 @@ static struct opt_params ropts = {
 				 { &ropts, R_RGSIZE },
 				 { NULL, LAST_CONFLICT } },
 		  .convert = true,
-		  .minval = 0,
+		  .minval = -1,
 		  .maxval = INT_MAX,
 		  .defaultval = 1,
 		},
@@ -1072,6 +1083,7 @@ struct cli_params {
 	int	blocksize;
 
 	char	*cfgfile;
+	bool	cfgfile_had_options;
 	char	*protofile;
 
 	enum fsprop_autofsck autofsck;
@@ -1654,10 +1666,12 @@ check_opt(
 		if (sp->seen)
 			respec(opts->name, opts->subopts, index);
 		sp->seen = true;
+		sp->from_file = parsing_cfgfile;
 	} else {
 		if (sp->str_seen)
 			respec(opts->name, opts->subopts, index);
 		sp->str_seen = true;
+		sp->from_file = parsing_cfgfile;
 	}
 
 	/* check for conflicts with the option */
@@ -1806,6 +1820,7 @@ set_data_concurrency(
 	/*
 	 * "nr_cpus" or "1" means set the concurrency level to the CPU count.
 	 * If this cannot be determined, fall back to the default AG geometry.
+	 * -1 means autodetect (use CPU count only on solid-state devices).
 	 */
 	if (!value || !strcmp(value, "nr_cpus"))
 		optnum = 1;
@@ -1814,6 +1829,8 @@ set_data_concurrency(
 
 	if (optnum == 1)
 		cli->data_concurrency = nr_cpus();
+	else if (optnum == -1)
+		cli->data_concurrency = -1;
 	else
 		cli->data_concurrency = optnum;
 }
@@ -1954,6 +1971,7 @@ set_log_concurrency(
 	/*
 	 * "nr_cpus" or 1 means set the concurrency level to the CPU count.  If
 	 * this cannot be determined, fall back to the default computation.
+	 * -1 means autodetect (use CPU count only on solid-state devices).
 	 */
 	if (!value || !strcmp(value, "nr_cpus"))
 		optnum = 1;
@@ -1962,6 +1980,8 @@ set_log_concurrency(
 
 	if (optnum == 1)
 		cli->log_concurrency = nr_cpus();
+	else if (optnum == -1)
+		cli->log_concurrency = -1;
 	else
 		cli->log_concurrency = optnum;
 }
@@ -2175,6 +2195,7 @@ set_rtvol_concurrency(
 	 * "nr_cpus" or "1" means set the concurrency level to the CPU count.
 	 * If this cannot be determined, fall back to the default rtgroup
 	 * geometry.
+	 * -1 means autodetect (use CPU count only on solid-state devices).
 	 */
 	if (!value || !strcmp(value, "nr_cpus"))
 		optnum = 1;
@@ -2183,6 +2204,8 @@ set_rtvol_concurrency(
 
 	if (optnum == 1)
 		cli->rtvol_concurrency = nr_cpus();
+	else if (optnum == -1)
+		cli->rtvol_concurrency = -1;
 	else
 		cli->rtvol_concurrency = optnum;
 }
@@ -2336,9 +2359,32 @@ parse_cfgopt(
 		if (!subopts[i])
 			break;
 		if (strcasecmp(name, subopts[i]) == 0) {
+			struct subopt_param	*sp = &sop->opts->subopt_params[i];
+			int			j;
+
+			/*
+			 * Command line options take precedence over config file
+			 * options. If this option or any option that conflicts
+			 * with it was already set from the command line, skip
+			 * the config file value silently.
+			 */
+			if ((sp->seen || sp->str_seen) && !sp->from_file)
+				return true;
+			for (j = 0; j < MAX_CONFLICTS; j++) {
+				struct _conflict	*con = &sp->conflicts[j];
+				struct subopt_param	*csp;
+
+				if (con->subopt == LAST_CONFLICT)
+					break;
+				csp = &con->opts->subopt_params[con->subopt];
+				if ((csp->seen || csp->str_seen) && !csp->from_file)
+					return true;
+			}
+
 			ret = (sop->parser)(sop->opts, i, value, cli);
 			if (ret)
 				goto invalid_opt;
+			cli->cfgfile_had_options = true;
 			return true;
 		}
 	}
@@ -5749,10 +5795,21 @@ cfgfile_parse(
 {
 	int			error;
 
-	if (!cli->cfgfile)
-		return;
+	bool	default_cfgfile = !cli->cfgfile;

mkfs/xfs_mkfs.c cfgfile_parse( · line 5808

Quoted 10 lines
mkfs/xfs_mkfs.c +10 -0
+
+	if (!cli->cfgfile) {
+		/*
+		 * No config file specified on the command line. Use the default
+		 * system-wide config file if it exists, otherwise do nothing.
+		 */
+		if (access(MKFS_DEFAULT_CFGFILE, F_OK) != 0)
+			return;
+		cli->cfgfile = MKFS_DEFAULT_CFGFILE;
+	}

This seems wrong to me. The defaults should always be parsed if the default file is present, whilst the CLI conf file should -overlay- the defaults like it currently does.

For example, the system might be set up for on, say 6.12-lts, so it's default is lts_6.12.conf. The user then has a custom config file for their cloud deployments that need some extra config (e.g. turns off rmapbt). i.e. they want all of the 6.12 configs except for one.

Instead of having to craft a config file for all the 6.12 defaults, they simple have "rmapbt=0" in a conf file and provide that on the CLI. Now they have a mkfs.xfs that defaults to 6.12-lts compatible filesystems, and when making their cloud images it simply applies a config overlay file via the command line.

When they upgrade all the systems to, say, 6.19-lts, they don't need to rewrite their custom overlay conf file - the system mkfs.xfs now defaults to 6.19-lts compatible filesystems, and their overlay conf file doesn't need changing. If they still need everything to use 6.12 comaptible filesystems, then they change the system default conf file, not their cloud-specific conf file.

mkfs/xfs_mkfs.c cfgfile_parse( · line 5815

Quoted 7 lines
mkfs/xfs_mkfs.c +2 -0
+	parsing_cfgfile = true;
 	error = ini_parse(cli->cfgfile, cfgfile_parse_ini, cli);
+	parsing_cfgfile = false;
 	if (error) {
 		if (error > 0) {
 			fprintf(stderr,

This is not how I envisiaged default config file setup to work for mkfs back when I rewrote all the parsing to be able to support config files. i.e.

commit 68344ba0f8cea778919e17958969b6c2459f890a

Author: Dave Chinner <[email protected]>
Date:   Wed Dec 6 17:14:27 2017 -0600

mkfs: introduce default configuration structure

mkfs has lots of options that require default values. Some of these are centralised, but others aren't. Introduce a new structure designed to hold default values for all the parameters that need defaults in one place.

This structure also provides a mechanism for providing mkfs defaults from a config file. This is not implemented in this series, but a comment is left where it is expected this functionality will hook in.

That comment is right at the start of the main() function:

/*
* TODO: Sourcing defaults from a config file
*
* Before anything else, see if there's a config file with different
* defaults. If a file exists in <package location>, read in the new
* default values and overwrite them in the &dft structure. This way the
* new defaults will apply before we parse the CLI, and the CLI will
* still be able to override them. When more than one source is
* implemented, emit a message to indicate where the defaults being
* used came from.
*
* printf(_("Default configuration sourced from %s\n"), dft.source); */

/* copy new defaults into CLI parsing structure */ memcpy(&cli.sb_feat, &dft.sb_feat, sizeof(cli.sb_feat)); memcpy(&cli.fsx, &dft.fsx, sizeof(cli.fsx));

What the new defaults file code should be doing is parsing the defaults file into a new "struct cli_params" structure, then using the options parsed from the file to initialise the CLI parsing structure. i.e. I had intended the above code to be replaced with something like:

	struct cli_params file_dfts { ....};

....

        /* Source defaults from a config file if it exists */
	file_dfts.config_file = MKFS_DEFAULT_CFGFILE;
        memcpy(&cli.sb_feat, &file_dfts.sb_feat, sizeof(cli.sb_feat));
        memcpy(&cli.fsx, &file_dfts.fsx, sizeof(cli.fsx));
	cfgfile_parse(&file_dfts);

	/* Set up defaults for CLI parsing */
	intialise_cli_params(&cli, &file_dfts);

.....

and initialise_cli_params() looked something like:

initialise_cli_params(

	*cli,
	*default)

{
/* copy feature fields into CLI parsing structure */ memcpy(&cli->sb_feat, &default->sb_feat, sizeof(cli->sb_feat)); memcpy(&cli->fsx, &default->fsx, sizeof(cli->fsx));

/*
* Check for parameters we shouldn't set in default conf
* files. e.g. device specific parameters like sector sizes,
* sunit/swidth, etc probably shouldn't be set in global
* default conf files. Warn and clear these parameters. */

/*
* Iterate the parameters found in the default config file
* and set them as initial values for the CLI parameter
* parsing. CLI parsing will overwrite these. */
<exercise left to the reader>
}

Now the CLI and cli conf file parsing should be able to be done pretty much as it is currently done. They should override the defaults as they currently do, and all existing scripts will behave as expected except for mkfs using different system-specified defaults.

-Dave.

Signature · 2 lines
Dave Chinner
[email protected]
#9 lore raw

Re: [RFC PATCH 1/1] xfsprogs: mkfs.xfs add default configuration file.

Quoted 7 lines
Quoted 2 lines

+[block]
+#size = 4096

Why do we need a config file that contains all the current defaults commented out? We don't do this for the command line conf files we ship, so this just seems like unnecessary maintenance overhead to me...

That might be unnecessary overhead and we might go just with the example file placed in /usr/share/

---

# xfs_admin -d /usr/share/xfsprogs/mkfs/lts_6.12.conf
And to reset mkfs behaviour back to built in defaults:
# xfs_admin -d clear
---
Quoted 36 lines
Quoted 28 lines
mkfs/mkfs.xfs.conf new file +14 -0
+
+[metadata]
+#crc = 1
+#finobt = 1
+#inobtcount = 1
+#rmapbt = 1
+#reflink = 1
+#bigtime = 1
+#metadir = 0
+#autofsck = 0
+
+[inode]
+#align = 1
+# The default value is 25% for filesystems under 1TB,#5% for 

filesystems under

mkfs/mkfs.xfs.conf new file +13 -0
+# 50TB and 1% for filesystems over 50TB.
+#
+#maxpct = 25
+#size = 512
+#perblock = 8
+#attr = 2
+#projid32bit = 1
+#sparse = 1
+#nrext64 = 1
+#exchange = 1
+
+[data]
+# -1 = autodetect (use CPU count only on solid-state devices), 0 = 
Quoted 2 lines

disabled
+#concurrency = -1

Not sure I like this way saying "autotune". This should match the existing conf file behaviour to override the default. i.e. 0 turns it off, 1 = autodetect, "nr_cpus" = autodetect", any other positive value is the minimum concurrency value.

Quoted 4 lines

IMO we should be exactly consistent with the CLI options here so that we don't need special parsing code just for the default options.

Noted, auto|autodetect is more descriptive than -1.

Quoted 137 lines

mkfs/xfs_mkfs.c cfgfile_parse( · line 5808

Quoted 9 lines
mkfs/xfs_mkfs.c +9 -0
+	if (!cli->cfgfile) {
+		/*
+		 * No config file specified on the command line. Use the default
+		 * system-wide config file if it exists, otherwise do nothing.
+		 */
+		if (access(MKFS_DEFAULT_CFGFILE, F_OK) != 0)
+			return;
+		cli->cfgfile = MKFS_DEFAULT_CFGFILE;
+	}

This seems wrong to me. The defaults should always be parsed if the default file is present, whilst the CLI conf file should -overlay- the defaults like it currently does.

For example, the system might be set up for on, say 6.12-lts, so it's default is lts_6.12.conf. The user then has a custom config file for their cloud deployments that need some extra config (e.g. turns off rmapbt). i.e. they want all of the 6.12 configs except for one.

Instead of having to craft a config file for all the 6.12 defaults, they simple have "rmapbt=0" in a conf file and provide that on the CLI. Now they have a mkfs.xfs that defaults to 6.12-lts compatible filesystems, and when making their cloud images it simply applies a config overlay file via the command line.

When they upgrade all the systems to, say, 6.19-lts, they don't need to rewrite their custom overlay conf file - the system mkfs.xfs now defaults to 6.19-lts compatible filesystems, and their overlay conf file doesn't need changing. If they still need everything to use 6.12 comaptible filesystems, then they change the system default conf file, not their cloud-specific conf file.

mkfs/xfs_mkfs.c cfgfile_parse( · line 5815

Quoted 7 lines
mkfs/xfs_mkfs.c +2 -0
+	parsing_cfgfile = true;
 	error = ini_parse(cli->cfgfile, cfgfile_parse_ini, cli);
+	parsing_cfgfile = false;
 	if (error) {
 		if (error > 0) {
 			fprintf(stderr,

This is not how I envisiaged default config file setup to work for mkfs back when I rewrote all the parsing to be able to support config files. i.e.

commit 68344ba0f8cea778919e17958969b6c2459f890a

Author: Dave Chinner <[email protected]>
Date:   Wed Dec 6 17:14:27 2017 -0600

mkfs: introduce default configuration structure

mkfs has lots of options that require default values. Some of these are centralised, but others aren't. Introduce a new structure designed to hold default values for all the parameters that need defaults in one place.

This structure also provides a mechanism for providing mkfs
defaults
from a config file. This is not implemented in this series, but a comment is left where it is expected this functionality will hook in.

That comment is right at the start of the main() function:

/*
* TODO: Sourcing defaults from a config file
*
* Before anything else, see if there's a config file with
different
* defaults. If a file exists in <package location>, read in
the new
* default values and overwrite them in the &dft structure.
This way the
* new defaults will apply before we parse the CLI, and the CLI
will
* still be able to override them. When more than one source is
* implemented, emit a message to indicate where the defaults
being
* used came from.
*
* printf(_("Default configuration sourced from %s\n"),

dft.source);
         */

        /* copy new defaults into CLI parsing structure */
        memcpy(&cli.sb_feat, &dft.sb_feat, sizeof(cli.sb_feat));
        memcpy(&cli.fsx, &dft.fsx, sizeof(cli.fsx));

What the new defaults file code should be doing is parsing the defaults file into a new "struct cli_params" structure, then using the options parsed from the file to initialise the CLI parsing structure. i.e. I had intended the above code to be replaced with something like:

	struct cli_params file_dfts { ....};

....

        /* Source defaults from a config file if it exists */
	file_dfts.config_file = MKFS_DEFAULT_CFGFILE;
        memcpy(&cli.sb_feat, &file_dfts.sb_feat, sizeof(cli.sb_feat));
        memcpy(&cli.fsx, &file_dfts.fsx, sizeof(cli.fsx));
	cfgfile_parse(&file_dfts);

	/* Set up defaults for CLI parsing */
	intialise_cli_params(&cli, &file_dfts);

.....

and initialise_cli_params() looked something like:

initialise_cli_params(

	*cli,
	*default)

{
/* copy feature fields into CLI parsing structure */ memcpy(&cli->sb_feat, &default->sb_feat, sizeof(cli->sb_feat)); memcpy(&cli->fsx, &default->fsx, sizeof(cli->fsx));

/*
* Check for parameters we shouldn't set in default conf
* files. e.g. device specific parameters like sector sizes,
* sunit/swidth, etc probably shouldn't be set in global
* default conf files. Warn and clear these parameters. */

/*
* Iterate the parameters found in the default config file
* and set them as initial values for the CLI parameter
* parsing. CLI parsing will overwrite these. */
<exercise left to the reader>
}

Quoted 7 lines

Now the CLI and cli conf file parsing should be able to be done pretty much as it is currently done. They should override the defaults as they currently do, and all existing scripts will behave as expected except for mkfs using different system-specified defaults.

-Dave.

Thanks for the example.
I should have read the comments more carefully, so if I get it right the logic should be: default.file < cli.options < cli.options.file?

Thanks!

Signature · 1 line
-lhe
#10 lore raw
In reply to Lukas Herbolt

[add ted to cc]

Quoted 5 lines · Lukas Herbolt

Hi all,
The default values cannot always satisfy everybody and users/packagers should be able to override defaults without the need of specifying configuration file on command line. It was previously discussed in here [1].

Hmm. A long time ago Ted asked for a slghtly different alteration to help him make running fstests easier. What he wanted iirc was a way to install a set of default mkfs settings from a config file, but still allow the settings to be overridden by cli arguments.

Apparently I never sent the patch because it doesn't exist in lore, so I suppose I should send it now. ;)

Quoted 4 lines

The proposed patch adds /etc/mkfs.xfs.conf where the default values can be set. It also lightly changes the default behavior for -l/-d/-r concurrency allowing now to be set to -1 to keep autodetection on SSDs.

I don't think I like magic negative numbers, but I'll save that for my reply to the actual patch.

--D

Quoted 16 lines

https://lore.kernel.org/linux-xfs/[email protected]/

Lukas Herbolt (1):
xfsprogs: mkfs.xfs add default configuration file.

 include/builddefs.in |  1 +
 mkfs/Makefile        |  3 ++
 mkfs/mkfs.xfs.conf   | 50 ++++++++++++++++++++++++++++++
 mkfs/xfs_mkfs.c      | 72 +++++++++++++++++++++++++++++++++++++++-----
 4 files changed, 119 insertions(+), 7 deletions(-)
 create mode 100644 mkfs/mkfs.xfs.conf
Signature · 1 line
2.54.0