Skip to contentSkip to conversation

kconfig: place git SHA1 in .config output if in SCM

PATCH In review Paul E. McKenney 3 participants 6 messages Mar 3, 2010 Kconfig

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 kconfig: place git SHA1 in .config output if in SCM

This patch appends the localversion string to the Linux kernel version. For example, in a git tree with uncommitted changes, the .config file might start as follows (but with leading hash marks):

	Automatically generated make config: don't edit
	Linux kernel version: 2.6.33-01836-g90a6501-dirty
	Mon Mar  1 17:05:59 2010

The "-01836-g90a6501-dirty" string is added by this patch.

This version of the patch incorporates feedback from Geert Uytterhoeven, Linus Torvalds, Frans Pop, and James Cloos:

o Fixed to work correctly with the "O=" Makefile argument and
the KBUILD_OUTPUT environment variable, so that .config files created in directories outside of the source tree are tagged correctly.

o Uses scripts/setlocalversion, which handles not only git, but
also mercurial and svn.

o Make the new behavior default-off, as scripts/setlocalversion
has significant latency. A new environment variable named "KBUILD_CONFIG_LOCALVERSION" must be set to enable the "-01836-g90a6501-dirty" style of string.

This is intended to address James Cloos's concern that this feature will slow down casual kernel builds.

It has been suggested that this string be output at boot and oops time. If there is general agreement, this will be the subject of a separate patch.

---

 confdata.c |   22 ++++++++++++++++++++--
 1 file changed, 20 insertions(+), 2 deletions(-)
scripts/kconfig/confdata.c +20 -2
index c4dec80..ea7e750 100644
@@ -399,10 +399,12 @@ int conf_read(const char *name)
 int conf_write(const char *name)
 {
 	FILE *out;
+	FILE *slv;
 	struct symbol *sym;
 	struct menu *menu;
 	const char *basename;
-	char dirname[128], tmpname[128], newname[128];
+	char dirname[128], tmpname[128], newname[128], localversion[128];
+	char cmdline[PATH_MAX * 2 + 128];
 	int type, l;
 	const char *str;
 	time_t now;
@@ -450,12 +452,28 @@ int conf_write(const char *name)
 	if (env && *env)
 		use_timestamp = 0;
 
+	localversion[0] = '\0';
+	if (getenv("KBUILD_CONFIG_LOCALVERSION")) {
+		env = getenv(SRCTREE);
+		if (env) {
+			sprintf(cmdline,
+				"%s/scripts/setlocalversion %s 2> /dev/null",
+				env, env);
+			slv = popen(cmdline, "r");
+			if (slv != NULL) {
+				fscanf(slv, " %127s ", localversion);
+				pclose(slv);
+			}
+		}
+	}
+
 	fprintf(out, _("#\n"
 		       "# Automatically generated make config: don't edit\n"
-		       "# Linux kernel version: %s\n"
+		       "# Linux kernel version: %s%s\n"
 		       "%s%s"
 		       "#\n"),
 		     sym_get_string_value(sym),
+		     localversion[0] != '\0' ? localversion : "",
 		     use_timestamp ? "# " : "",
 		     use_timestamp ? ctime(&now) : "");
 
#1 lore raw

scripts/kconfig/confdata.c int conf_write(const char *name) · line 462

Quoted 6 lines · Paul E. McKenney
scripts/kconfig/confdata.c +6 -0
+		env = getenv(SRCTREE);
+		if (env) {
+			sprintf(cmdline,
+				"%s/scripts/setlocalversion %s 2> /dev/null",
+				env, env);
+			slv = popen(cmdline, "r");

I suspect this does various bad things if there are spaces or special characters in $SRCTREE.

It would be a lot safer to uses fork/execve rather than something that interprets a shell command line.

Of course, I didn't check that all our old users of SRCTREE are safe either, but at least docproc.c (the one I _did_ check) uses 'execvp()' and 'fopen()' that both take real filenames, not a shell string.

Linus

#2 lore raw
Quoted 13 lines · Linus Torvalds

scripts/kconfig/confdata.c int conf_write(const char *name) · line 462

Quoted 6 lines · Paul E. McKenney
scripts/kconfig/confdata.c +6 -0
+		env = getenv(SRCTREE);
+		if (env) {
+			sprintf(cmdline,
+				"%s/scripts/setlocalversion %s 2> /dev/null",
+				env, env);
+			slv = popen(cmdline, "r");

I suspect this does various bad things if there are spaces or special characters in $SRCTREE.

Quoted 6 lines

It would be a lot safer to uses fork/execve rather than something that interprets a shell command line.

Of course, I didn't check that all our old users of SRCTREE are safe either, but at least docproc.c (the one I _did_ check) uses 'execvp()' and 'fopen()' that both take real filenames, not a shell string.

Well, we certainly don't want or need bash's "$", "``", and other interpretations in this case. I will update and send out a new patch.

Hmmm... It has been one good long time since I have used pipe(), dup2(), exec*(), and friends. In happy contrast to last time, some of the man pages now seem to have nice examples. ;-)

Thanx, Paul

#3 lore raw

Doesn't this patch simply duplicate CONFIG_LOCALVERSION_AUTO's functionality, now that it calls scripts/setlocalversion?

I'd still like to get the hash and only elide the dirty check.

-JimC

Signature · 1 line
James Cloos <[email protected]>         OpenPGP: 1024D/ED7DAEA6
#4 lore raw
Quoted 4 lines · James Cloos

Doesn't this patch simply duplicate CONFIG_LOCALVERSION_AUTO's functionality, now that it calls scripts/setlocalversion?

I'd still like to get the hash and only elide the dirty check.

Would you be OK with having to set an environment variable to suppress the dirty check?

Thanx, Paul

#5 lore raw
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line

"Paul" == Paul E McKenney <[email protected]> writes:

JimC> Doesn't this patch simply duplicate CONFIG_LOCALVERSION_AUTO's JimC> functionality, now that it calls scripts/setlocalversion?

JimC> I'd still like to get the hash and only elide the dirty check.

Paul> Would you be OK with having to set an environment variable to suppress Paul> the dirty check?

Yes, most definitely OK.

-JimC

Signature · 1 line
James Cloos <[email protected]>         OpenPGP: 1024D/ED7DAEA6