Skip to contentSkip to conversation

kexec: do KEXEC_FILE_LOAD and fallback to KEXEC_LOAD if not supported.

PATCH In review Michal Suchanek 4 participants 15 messages Feb 22, 2018 – Feb 27, 2018

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 kexec: do KEXEC_FILE_LOAD and fallback to KEXEC_LOAD if not supported.

The new KEXEC_FILE_LOAD is preferred in the case the platform supports it because it allows kexec in locked down secure boot mode.

However, some platforms do not support it so fall back to the old syscall there.

Also provide an option to call the old syscall in case the new syscall fails with other reason than ENOSYS.

Also document the options.

---
 kexec/kexec.8 |  9 +++++++++
 kexec/kexec.c | 41 +++++++++++++++--------------------------
 kexec/kexec.h |  2 ++
 3 files changed, 26 insertions(+), 26 deletions(-)
kexec/kexec.8 +9 -0
index e0131b4ea827..7e4df723251d 100644
@@ -144,6 +144,15 @@ Load the new kernel for use on panic.
 Specify that the new kernel is of this
 .I type.
 .TP
+.BI \-s\ (\-\-kexec-file-syscall)
+Specify that the new KEXEC_FILE_LOAD syscall should be used exclusively.
+Otherwise KEXEC_FILE_LOAD is tried and when not supported KEXEC_LOAD is used.
+.I type.
+.TP
+.BI \-c\ (\-\-kexec-syscall)
+Specify that the old KEXEC_LOAD syscall should be used exclusively.
+.I type.
+.TP
 .B \-u\ (\-\-unload)
 Unload the current
 .B kexec
kexec/kexec.c +15 -26

89 lines not on this page · raw message

kexec/kexec.h +2 -0
index 26225d2c002a..7abcec796cae 100644
@@ -219,6 +219,7 @@ extern int file_types;
 #define OPT_TYPE		't'
 #define OPT_PANIC		'p'
 #define OPT_KEXEC_FILE_SYSCALL	's'
+#define OPT_KEXEC_SYSCALL	'c'
 #define OPT_STATUS		'S'
 #define OPT_MEM_MIN             256
 #define OPT_MEM_MAX             257
@@ -246,6 +247,7 @@ extern int file_types;
 	{ "mem-max",		1, 0, OPT_MEM_MAX }, \
 	{ "reuseinitrd",	0, 0, OPT_REUSE_INITRD }, \
 	{ "kexec-file-syscall",	0, 0, OPT_KEXEC_FILE_SYSCALL }, \
+	{ "kexec-syscall",	0, 0, OPT_KEXEC_SYSCALL }, \
 	{ "debug",		0, 0, OPT_DEBUG }, \
 	{ "status",		0, 0, OPT_STATUS }, \
 	{ "print-ckr-size",     0, 0, OPT_PRINT_CKR_SIZE }, \
Signature · 7 lines
2.13.6


_______________________________________________
kexec mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/kexec
#1 lore raw

Hi Michal,

Quoted 5 lines · Michal Suchanek

The new KEXEC_FILE_LOAD is preferred in the case the platform supports it because it allows kexec in locked down secure boot mode.

However, some platforms do not support it so fall back to the old syscall there.

I didn't read code change, just from patch log, I tend to not agree. There are two options KEXEC_FILE_LOAD and KEXEC_LOAD, some platforms do not support, why does some platforms not choose KEXEC_LOAD, the working one? Why bother to make change in code? I believe there's returned message telling if KEXEC_FILE_LOAD works or not.

Thanks
Baoquan

Quoted 161 lines

Also provide an option to call the old syscall in case the new syscall fails with other reason than ENOSYS.

Also document the options.

---
 kexec/kexec.8 |  9 +++++++++
 kexec/kexec.c | 41 +++++++++++++++--------------------------
 kexec/kexec.h |  2 ++
 3 files changed, 26 insertions(+), 26 deletions(-)
kexec/kexec.8 +9 -0
index e0131b4ea827..7e4df723251d 100644
@@ -144,6 +144,15 @@ Load the new kernel for use on panic.
 Specify that the new kernel is of this
 .I type.
 .TP
+.BI \-s\ (\-\-kexec-file-syscall)
+Specify that the new KEXEC_FILE_LOAD syscall should be used exclusively.
+Otherwise KEXEC_FILE_LOAD is tried and when not supported KEXEC_LOAD is used.
+.I type.
+.TP
+.BI \-c\ (\-\-kexec-syscall)
+Specify that the old KEXEC_LOAD syscall should be used exclusively.
+.I type.
+.TP
 .B \-u\ (\-\-unload)
 Unload the current
 .B kexec
kexec/kexec.c +15 -26
index cfd837c1b6bb..25328c02b508 100644
@@ -1166,7 +1166,7 @@ static int do_kexec_file_load(int fileind, int argc, char **argv,
 
 	if (!is_kexec_file_load_implemented()) {
 		fprintf(stderr, "syscall kexec_file_load not available.\n");
-		return -1;
+		return -ENOSYS;
 	}
 
 	if (argc - fileind <= 0) {
@@ -1243,6 +1243,7 @@ int main(int argc, char *argv[])
 	int do_unload = 0;
 	int do_reuse_initrd = 0;
 	int do_kexec_file_syscall = 0;
+	int do_kexec_syscall = 0;
 	int do_status = 0;
 	void *entry = 0;
 	char *type = 0;
@@ -1256,19 +1257,6 @@ int main(int argc, char *argv[])
 	};
 	static const char short_options[] = KEXEC_ALL_OPT_STR;
 
-	/*
-	 * First check if --use-kexec-file-syscall is set. That changes lot of
-	 * things
-	 */
-	while ((opt = getopt_long(argc, argv, short_options,
-				  options, 0)) != -1) {
-		switch(opt) {
-		case OPT_KEXEC_FILE_SYSCALL:
-			do_kexec_file_syscall = 1;
-			break;
-		}
-	}
-
 	/* Reset getopt for the next pass. */
 	opterr = 1;
 	optind = 1;
@@ -1310,8 +1298,7 @@ int main(int argc, char *argv[])
 			do_shutdown = 0;
 			do_sync = 0;
 			do_unload = 1;
-			if (do_kexec_file_syscall)
-				kexec_file_flags |= KEXEC_FILE_UNLOAD;
+			kexec_file_flags |= KEXEC_FILE_UNLOAD;
 			break;
 		case OPT_EXEC:
 			do_load = 0;
@@ -1354,11 +1341,8 @@ int main(int argc, char *argv[])
 			do_exec = 0;
 			do_shutdown = 0;
 			do_sync = 0;
-			if (do_kexec_file_syscall)
-				kexec_file_flags |= KEXEC_FILE_ON_CRASH;
-			else
-				kexec_flags = KEXEC_ON_CRASH;
-			break;
+			kexec_file_flags |= KEXEC_FILE_ON_CRASH;
+			kexec_flags = KEXEC_ON_CRASH;
 		case OPT_MEM_MIN:
 			mem_min = strtoul(optarg, &endptr, 0);
 			if (*endptr) {
@@ -1383,7 +1367,12 @@ int main(int argc, char *argv[])
 			do_reuse_initrd = 1;
 			break;
 		case OPT_KEXEC_FILE_SYSCALL:
-			/* We already parsed it. Nothing to do. */
+			do_kexec_file_syscall = 1;
+			do_kexec_syscall = 0;
+			break;
+		case OPT_KEXEC_SYSCALL:
+			do_kexec_file_syscall = 0;
+			do_kexec_syscall = 1;
 			break;
 		case OPT_STATUS:
 			do_status = 1;
@@ -1456,16 +1445,16 @@ int main(int argc, char *argv[])
 		result = k_status(kexec_flags);
 	}
 	if (do_unload) {
-		if (do_kexec_file_syscall)
+		if (!do_kexec_syscall)
 			result = kexec_file_unload(kexec_file_flags);
-		else
+		if ((result == -ENOSYS) || !do_kexec_file_syscall)
 			result = k_unload(kexec_flags);
 	}
 	if (do_load && (result == 0)) {
-		if (do_kexec_file_syscall)
+		if (!do_kexec_syscall)
 			result = do_kexec_file_load(fileind, argc, argv,
 						 kexec_file_flags);
-		else
+		if ((result == -ENOSYS) || !do_kexec_file_syscall)
 			result = my_load(type, fileind, argc, argv,
 						kexec_flags, entry);
 	}
kexec/kexec.h +2 -0
index 26225d2c002a..7abcec796cae 100644
@@ -219,6 +219,7 @@ extern int file_types;
 #define OPT_TYPE		't'
 #define OPT_PANIC		'p'
 #define OPT_KEXEC_FILE_SYSCALL	's'
+#define OPT_KEXEC_SYSCALL	'c'
 #define OPT_STATUS		'S'
 #define OPT_MEM_MIN             256
 #define OPT_MEM_MAX             257
@@ -246,6 +247,7 @@ extern int file_types;
 	{ "mem-max",		1, 0, OPT_MEM_MAX }, \
 	{ "reuseinitrd",	0, 0, OPT_REUSE_INITRD }, \
 	{ "kexec-file-syscall",	0, 0, OPT_KEXEC_FILE_SYSCALL }, \
+	{ "kexec-syscall",	0, 0, OPT_KEXEC_SYSCALL }, \
 	{ "debug",		0, 0, OPT_DEBUG }, \
 	{ "status",		0, 0, OPT_STATUS }, \
 	{ "print-ckr-size",     0, 0, OPT_PRINT_CKR_SIZE }, \
Signature · 7 lines
2.13.6


_______________________________________________
kexec mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/kexec

_______________________________________________
kexec mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/kexec

#2 lore raw

Hi Baoquan,

Quoted 6 lines · On Fri, 23 Feb 2018 07:20:43 +0800 Baoquan He <[email protected]> wrote:

Hi Michal,

Quoted 3 lines · Michal Suchanek

The new KEXEC_FILE_LOAD is preferred in the case the platform supports it because it allows kexec in locked down secure boot mode.

Quoted 8 lines
Quoted 2 lines

However, some platforms do not support it so fall back to the old syscall there.

I didn't read code change, just from patch log, I tend to not agree. There are two options KEXEC_FILE_LOAD and KEXEC_LOAD, some platforms do not support, why does some platforms not choose KEXEC_LOAD, the working one? Why bother to make change in code? I believe there's returned message telling if KEXEC_FILE_LOAD works or not.

Well... let me give a bit of background. As you have probably noticed, this syscall was originally available only for x86_64, but more and more architectures are also adding it now.

Next, kexec is actually called by a script (which locates a suitable kernel and initrd, constructs the kernel command line, etc.). The script must either:

A. know somehow if the currently running kernel implements

     kexec_file_load(2), or

B. try one method first, and if it fails, retry with the other.

I agree that kexec(1) should probably allow the user to force a specific method, but I don't see the benefit of implementing fallback in an external script and not in kexec-tools itself.

OTOH if you want to push the fallback logic out of kexec-tools, then I would like to get better diagnostic at least. Letting my script parse kexec output is, um, suboptimal.

Petr T

Quoted 159 lines

Thanks
Baoquan

Quoted 157 lines

Also provide an option to call the old syscall in case the new syscall fails with other reason than ENOSYS.

Also document the options.

---
 kexec/kexec.8 |  9 +++++++++
 kexec/kexec.c | 41 +++++++++++++++--------------------------
 kexec/kexec.h |  2 ++
 3 files changed, 26 insertions(+), 26 deletions(-)
kexec/kexec.8 +9 -0
index e0131b4ea827..7e4df723251d 100644
@@ -144,6 +144,15 @@ Load the new kernel for use on panic.
 Specify that the new kernel is of this
 .I type.
 .TP
+.BI \-s\ (\-\-kexec-file-syscall)
+Specify that the new KEXEC_FILE_LOAD syscall should be used exclusively.
+Otherwise KEXEC_FILE_LOAD is tried and when not supported KEXEC_LOAD is used.
+.I type.
+.TP
+.BI \-c\ (\-\-kexec-syscall)
+Specify that the old KEXEC_LOAD syscall should be used exclusively.
+.I type.
+.TP
 .B \-u\ (\-\-unload)
 Unload the current
 .B kexec
kexec/kexec.c +15 -26
index cfd837c1b6bb..25328c02b508 100644
@@ -1166,7 +1166,7 @@ static int do_kexec_file_load(int fileind, int argc, char **argv,
 
 	if (!is_kexec_file_load_implemented()) {
 		fprintf(stderr, "syscall kexec_file_load not available.\n");
-		return -1;
+		return -ENOSYS;
 	}
 
 	if (argc - fileind <= 0) {
@@ -1243,6 +1243,7 @@ int main(int argc, char *argv[])
 	int do_unload = 0;
 	int do_reuse_initrd = 0;
 	int do_kexec_file_syscall = 0;
+	int do_kexec_syscall = 0;
 	int do_status = 0;
 	void *entry = 0;
 	char *type = 0;
@@ -1256,19 +1257,6 @@ int main(int argc, char *argv[])
 	};
 	static const char short_options[] = KEXEC_ALL_OPT_STR;
 
-	/*
-	 * First check if --use-kexec-file-syscall is set. That changes lot of
-	 * things
-	 */
-	while ((opt = getopt_long(argc, argv, short_options,
-				  options, 0)) != -1) {
-		switch(opt) {
-		case OPT_KEXEC_FILE_SYSCALL:
-			do_kexec_file_syscall = 1;
-			break;
-		}
-	}
-
 	/* Reset getopt for the next pass. */
 	opterr = 1;
 	optind = 1;
@@ -1310,8 +1298,7 @@ int main(int argc, char *argv[])
 			do_shutdown = 0;
 			do_sync = 0;
 			do_unload = 1;
-			if (do_kexec_file_syscall)
-				kexec_file_flags |= KEXEC_FILE_UNLOAD;
+			kexec_file_flags |= KEXEC_FILE_UNLOAD;
 			break;
 		case OPT_EXEC:
 			do_load = 0;
@@ -1354,11 +1341,8 @@ int main(int argc, char *argv[])
 			do_exec = 0;
 			do_shutdown = 0;
 			do_sync = 0;
-			if (do_kexec_file_syscall)
-				kexec_file_flags |= KEXEC_FILE_ON_CRASH;
-			else
-				kexec_flags = KEXEC_ON_CRASH;
-			break;
+			kexec_file_flags |= KEXEC_FILE_ON_CRASH;
+			kexec_flags = KEXEC_ON_CRASH;
 		case OPT_MEM_MIN:
 			mem_min = strtoul(optarg, &endptr, 0);
 			if (*endptr) {
@@ -1383,7 +1367,12 @@ int main(int argc, char *argv[])
 			do_reuse_initrd = 1;
 			break;
 		case OPT_KEXEC_FILE_SYSCALL:
-			/* We already parsed it. Nothing to do. */
+			do_kexec_file_syscall = 1;
+			do_kexec_syscall = 0;
+			break;
+		case OPT_KEXEC_SYSCALL:
+			do_kexec_file_syscall = 0;
+			do_kexec_syscall = 1;
 			break;
 		case OPT_STATUS:
 			do_status = 1;
@@ -1456,16 +1445,16 @@ int main(int argc, char *argv[])
 		result = k_status(kexec_flags);
 	}
 	if (do_unload) {
-		if (do_kexec_file_syscall)
+		if (!do_kexec_syscall)
 			result = kexec_file_unload(kexec_file_flags);
-		else
+		if ((result == -ENOSYS) || !do_kexec_file_syscall)
 			result = k_unload(kexec_flags);
 	}
 	if (do_load && (result == 0)) {
-		if (do_kexec_file_syscall)
+		if (!do_kexec_syscall)
 			result = do_kexec_file_load(fileind, argc, argv,
 						 kexec_file_flags);
-		else
+		if ((result == -ENOSYS) || !do_kexec_file_syscall)
 			result = my_load(type, fileind, argc, argv,
 						kexec_flags, entry);
 	}
kexec/kexec.h +2 -0
index 26225d2c002a..7abcec796cae 100644
@@ -219,6 +219,7 @@ extern int file_types;
 #define OPT_TYPE		't'
 #define OPT_PANIC		'p'
 #define OPT_KEXEC_FILE_SYSCALL	's'
+#define OPT_KEXEC_SYSCALL	'c'
 #define OPT_STATUS		'S'
 #define OPT_MEM_MIN             256
 #define OPT_MEM_MAX             257
@@ -246,6 +247,7 @@ extern int file_types;
 	{ "mem-max",		1, 0, OPT_MEM_MAX }, \
 	{ "reuseinitrd",	0, 0, OPT_REUSE_INITRD }, \
 	{ "kexec-file-syscall",	0, 0, OPT_KEXEC_FILE_SYSCALL }, \
+	{ "kexec-syscall",	0, 0, OPT_KEXEC_SYSCALL }, \
 	{ "debug",		0, 0, OPT_DEBUG }, \
 	{ "status",		0, 0, OPT_STATUS }, \
 	{ "print-ckr-size",     0, 0, OPT_PRINT_CKR_SIZE }, \
Signature · 1 line
2.13.6
Quoted 9 lines
Quoted 4 lines

_______________________________________________
kexec mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/kexec

_______________________________________________
kexec mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/kexec

_______________________________________________
kexec mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/kexec

#3 lore raw
Quoted 33 lines · Petr Tesarik

Hi Baoquan,

Quoted 6 lines · On Fri, 23 Feb 2018 07:20:43 +0800 Baoquan He <[email protected]> wrote:

Hi Michal,

Quoted 3 lines · Michal Suchanek

The new KEXEC_FILE_LOAD is preferred in the case the platform supports it because it allows kexec in locked down secure boot mode.

Quoted 8 lines
Quoted 2 lines

However, some platforms do not support it so fall back to the old syscall there.

I didn't read code change, just from patch log, I tend to not agree. There are two options KEXEC_FILE_LOAD and KEXEC_LOAD, some platforms do not support, why does some platforms not choose KEXEC_LOAD, the working one? Why bother to make change in code? I believe there's returned message telling if KEXEC_FILE_LOAD works or not.

Well... let me give a bit of background. As you have probably noticed, this syscall was originally available only for x86_64, but more and more architectures are also adding it now.

Next, kexec is actually called by a script (which locates a suitable kernel and initrd, constructs the kernel command line, etc.). The script must either:

A. know somehow if the currently running kernel implements

     kexec_file_load(2), or

B. try one method first, and if it fails, retry with the other.

Quoted 7 lines

I agree that kexec(1) should probably allow the user to force a specific method, but I don't see the benefit of implementing fallback in an external script and not in kexec-tools itself.

OTOH if you want to push the fallback logic out of kexec-tools, then I would like to get better diagnostic at least. Letting my script parse kexec output is, um, suboptimal.

In Fedora/RHEL we use this in scripts by checking the arch first, for distribution it is enough? There are also some other arch dependent options in kexec-tools, there is no way to just use same for every different platform without checking in scripts.

If your scripts is not for a distribution, I agree that it is indeed a problem.

Quoted 173 lines

Petr T

Quoted 159 lines

Thanks
Baoquan

Quoted 157 lines

Also provide an option to call the old syscall in case the new syscall fails with other reason than ENOSYS.

Also document the options.

---
 kexec/kexec.8 |  9 +++++++++
 kexec/kexec.c | 41 +++++++++++++++--------------------------
 kexec/kexec.h |  2 ++
 3 files changed, 26 insertions(+), 26 deletions(-)
kexec/kexec.8 +9 -0
index e0131b4ea827..7e4df723251d 100644
@@ -144,6 +144,15 @@ Load the new kernel for use on panic.
 Specify that the new kernel is of this
 .I type.
 .TP
+.BI \-s\ (\-\-kexec-file-syscall)
+Specify that the new KEXEC_FILE_LOAD syscall should be used exclusively.
+Otherwise KEXEC_FILE_LOAD is tried and when not supported KEXEC_LOAD is used.
+.I type.
+.TP
+.BI \-c\ (\-\-kexec-syscall)
+Specify that the old KEXEC_LOAD syscall should be used exclusively.
+.I type.
+.TP
 .B \-u\ (\-\-unload)
 Unload the current
 .B kexec
kexec/kexec.c +15 -26
index cfd837c1b6bb..25328c02b508 100644
@@ -1166,7 +1166,7 @@ static int do_kexec_file_load(int fileind, int argc, char **argv,
 
 	if (!is_kexec_file_load_implemented()) {
 		fprintf(stderr, "syscall kexec_file_load not available.\n");
-		return -1;
+		return -ENOSYS;
 	}
 
 	if (argc - fileind <= 0) {
@@ -1243,6 +1243,7 @@ int main(int argc, char *argv[])
 	int do_unload = 0;
 	int do_reuse_initrd = 0;
 	int do_kexec_file_syscall = 0;
+	int do_kexec_syscall = 0;
 	int do_status = 0;
 	void *entry = 0;
 	char *type = 0;
@@ -1256,19 +1257,6 @@ int main(int argc, char *argv[])
 	};
 	static const char short_options[] = KEXEC_ALL_OPT_STR;
 
-	/*
-	 * First check if --use-kexec-file-syscall is set. That changes lot of
-	 * things
-	 */
-	while ((opt = getopt_long(argc, argv, short_options,
-				  options, 0)) != -1) {
-		switch(opt) {
-		case OPT_KEXEC_FILE_SYSCALL:
-			do_kexec_file_syscall = 1;
-			break;
-		}
-	}
-
 	/* Reset getopt for the next pass. */
 	opterr = 1;
 	optind = 1;
@@ -1310,8 +1298,7 @@ int main(int argc, char *argv[])
 			do_shutdown = 0;
 			do_sync = 0;
 			do_unload = 1;
-			if (do_kexec_file_syscall)
-				kexec_file_flags |= KEXEC_FILE_UNLOAD;
+			kexec_file_flags |= KEXEC_FILE_UNLOAD;
 			break;
 		case OPT_EXEC:
 			do_load = 0;
@@ -1354,11 +1341,8 @@ int main(int argc, char *argv[])
 			do_exec = 0;
 			do_shutdown = 0;
 			do_sync = 0;
-			if (do_kexec_file_syscall)
-				kexec_file_flags |= KEXEC_FILE_ON_CRASH;
-			else
-				kexec_flags = KEXEC_ON_CRASH;
-			break;
+			kexec_file_flags |= KEXEC_FILE_ON_CRASH;
+			kexec_flags = KEXEC_ON_CRASH;
 		case OPT_MEM_MIN:
 			mem_min = strtoul(optarg, &endptr, 0);
 			if (*endptr) {
@@ -1383,7 +1367,12 @@ int main(int argc, char *argv[])
 			do_reuse_initrd = 1;
 			break;
 		case OPT_KEXEC_FILE_SYSCALL:
-			/* We already parsed it. Nothing to do. */
+			do_kexec_file_syscall = 1;
+			do_kexec_syscall = 0;
+			break;
+		case OPT_KEXEC_SYSCALL:
+			do_kexec_file_syscall = 0;
+			do_kexec_syscall = 1;
 			break;
 		case OPT_STATUS:
 			do_status = 1;
@@ -1456,16 +1445,16 @@ int main(int argc, char *argv[])
 		result = k_status(kexec_flags);
 	}
 	if (do_unload) {
-		if (do_kexec_file_syscall)
+		if (!do_kexec_syscall)
 			result = kexec_file_unload(kexec_file_flags);
-		else
+		if ((result == -ENOSYS) || !do_kexec_file_syscall)
 			result = k_unload(kexec_flags);
 	}
 	if (do_load && (result == 0)) {
-		if (do_kexec_file_syscall)
+		if (!do_kexec_syscall)
 			result = do_kexec_file_load(fileind, argc, argv,
 						 kexec_file_flags);
-		else
+		if ((result == -ENOSYS) || !do_kexec_file_syscall)
 			result = my_load(type, fileind, argc, argv,
 						kexec_flags, entry);
 	}
kexec/kexec.h +2 -0
index 26225d2c002a..7abcec796cae 100644
@@ -219,6 +219,7 @@ extern int file_types;
 #define OPT_TYPE		't'
 #define OPT_PANIC		'p'
 #define OPT_KEXEC_FILE_SYSCALL	's'
+#define OPT_KEXEC_SYSCALL	'c'
 #define OPT_STATUS		'S'
 #define OPT_MEM_MIN             256
 #define OPT_MEM_MAX             257
@@ -246,6 +247,7 @@ extern int file_types;
 	{ "mem-max",		1, 0, OPT_MEM_MAX }, \
 	{ "reuseinitrd",	0, 0, OPT_REUSE_INITRD }, \
 	{ "kexec-file-syscall",	0, 0, OPT_KEXEC_FILE_SYSCALL }, \
+	{ "kexec-syscall",	0, 0, OPT_KEXEC_SYSCALL }, \
 	{ "debug",		0, 0, OPT_DEBUG }, \
 	{ "status",		0, 0, OPT_STATUS }, \
 	{ "print-ckr-size",     0, 0, OPT_PRINT_CKR_SIZE }, \
Signature · 1 line
2.13.6
Quoted 9 lines
Quoted 4 lines

_______________________________________________
kexec mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/kexec

_______________________________________________
kexec mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/kexec

Quoted 4 lines

_______________________________________________
kexec mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/kexec

Thanks
Dave

_______________________________________________
kexec mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/kexec

#4 lore raw
Quoted 38 lines · On Sat, 24 Feb 2018 09:43:42 +0800 Dave Young <[email protected]> wrote:
Quoted 37 lines · Petr Tesarik

Hi Baoquan,

Quoted 6 lines · On Fri, 23 Feb 2018 07:20:43 +0800 Baoquan He <[email protected]> wrote:

Hi Michal,

Quoted 3 lines · Michal Suchanek

The new KEXEC_FILE_LOAD is preferred in the case the platform supports it because it allows kexec in locked down secure boot mode.

Quoted 8 lines
Quoted 2 lines

However, some platforms do not support it so fall back to the old syscall there.

I didn't read code change, just from patch log, I tend to not agree. There are two options KEXEC_FILE_LOAD and KEXEC_LOAD, some platforms do not support, why does some platforms not choose KEXEC_LOAD, the working one? Why bother to make change in code? I believe there's returned message telling if KEXEC_FILE_LOAD works or not.

Well... let me give a bit of background. As you have probably noticed, this syscall was originally available only for x86_64, but more and more architectures are also adding it now.

Next, kexec is actually called by a script (which locates a suitable kernel and initrd, constructs the kernel command line, etc.). The script must either:

A. know somehow if the currently running kernel implements

     kexec_file_load(2), or

B. try one method first, and if it fails, retry with the other.

I agree that kexec(1) should probably allow the user to force a specific method, but I don't see the benefit of implementing fallback in an external script and not in kexec-tools itself.

Quoted 6 lines
Quoted 3 lines

OTOH if you want to push the fallback logic out of kexec-tools, then I would like to get better diagnostic at least. Letting my script parse kexec output is, um, suboptimal.

In Fedora/RHEL we use this in scripts by checking the arch first, for distribution it is enough?

No.

First, you would also have to check the kernel version (and maintain an ugly mapping of which kernel version introduced kexec_file_load on which architecture).

Second, it's not just the architecture. kexec_load(2) will fail if SecureBoot is active. OTOH kexec_file_load(2) will fail if the kernel is not signed. For kernel hackers who don't use SecureBoot, signing self-built kernels is just overkill. So, you should also check the state of SecureBoot, possibly also whether the kernel image is signed with a valid key, repeating a bit too much of the kernel logic, and quite likely introducing subtle differences...

Petr T

Quoted 179 lines

There are also some other arch dependent
options in kexec-tools, there is no way to just use same for every different platform without checking in scripts.

If your scripts is not for a distribution, I agree that it is indeed a problem.

Quoted 173 lines

Petr T

Quoted 168 lines

Thanks
Baoquan

Quoted 161 lines

Also provide an option to call the old syscall in case the new syscall fails with other reason than ENOSYS.

Also document the options.

---
 kexec/kexec.8 |  9 +++++++++
 kexec/kexec.c | 41 +++++++++++++++--------------------------
 kexec/kexec.h |  2 ++
 3 files changed, 26 insertions(+), 26 deletions(-)
kexec/kexec.8 +9 -0
index e0131b4ea827..7e4df723251d 100644
@@ -144,6 +144,15 @@ Load the new kernel for use on panic.
 Specify that the new kernel is of this
 .I type.
 .TP
+.BI \-s\ (\-\-kexec-file-syscall)
+Specify that the new KEXEC_FILE_LOAD syscall should be used exclusively.
+Otherwise KEXEC_FILE_LOAD is tried and when not supported KEXEC_LOAD is used.
+.I type.
+.TP
+.BI \-c\ (\-\-kexec-syscall)
+Specify that the old KEXEC_LOAD syscall should be used exclusively.
+.I type.
+.TP
 .B \-u\ (\-\-unload)
 Unload the current
 .B kexec
kexec/kexec.c +15 -26
index cfd837c1b6bb..25328c02b508 100644
@@ -1166,7 +1166,7 @@ static int do_kexec_file_load(int fileind, int argc, char **argv,
 
 	if (!is_kexec_file_load_implemented()) {
 		fprintf(stderr, "syscall kexec_file_load not available.\n");
-		return -1;
+		return -ENOSYS;
 	}
 
 	if (argc - fileind <= 0) {
@@ -1243,6 +1243,7 @@ int main(int argc, char *argv[])
 	int do_unload = 0;
 	int do_reuse_initrd = 0;
 	int do_kexec_file_syscall = 0;
+	int do_kexec_syscall = 0;
 	int do_status = 0;
 	void *entry = 0;
 	char *type = 0;
@@ -1256,19 +1257,6 @@ int main(int argc, char *argv[])
 	};
 	static const char short_options[] = KEXEC_ALL_OPT_STR;
 
-	/*
-	 * First check if --use-kexec-file-syscall is set. That changes lot of
-	 * things
-	 */
-	while ((opt = getopt_long(argc, argv, short_options,
-				  options, 0)) != -1) {
-		switch(opt) {
-		case OPT_KEXEC_FILE_SYSCALL:
-			do_kexec_file_syscall = 1;
-			break;
-		}
-	}
-
 	/* Reset getopt for the next pass. */
 	opterr = 1;
 	optind = 1;
@@ -1310,8 +1298,7 @@ int main(int argc, char *argv[])
 			do_shutdown = 0;
 			do_sync = 0;
 			do_unload = 1;
-			if (do_kexec_file_syscall)
-				kexec_file_flags |= KEXEC_FILE_UNLOAD;
+			kexec_file_flags |= KEXEC_FILE_UNLOAD;
 			break;
 		case OPT_EXEC:
 			do_load = 0;
@@ -1354,11 +1341,8 @@ int main(int argc, char *argv[])
 			do_exec = 0;
 			do_shutdown = 0;
 			do_sync = 0;
-			if (do_kexec_file_syscall)
-				kexec_file_flags |= KEXEC_FILE_ON_CRASH;
-			else
-				kexec_flags = KEXEC_ON_CRASH;
-			break;
+			kexec_file_flags |= KEXEC_FILE_ON_CRASH;
+			kexec_flags = KEXEC_ON_CRASH;
 		case OPT_MEM_MIN:
 			mem_min = strtoul(optarg, &endptr, 0);
 			if (*endptr) {
@@ -1383,7 +1367,12 @@ int main(int argc, char *argv[])
 			do_reuse_initrd = 1;
 			break;
 		case OPT_KEXEC_FILE_SYSCALL:
-			/* We already parsed it. Nothing to do. */
+			do_kexec_file_syscall = 1;
+			do_kexec_syscall = 0;
+			break;
+		case OPT_KEXEC_SYSCALL:
+			do_kexec_file_syscall = 0;
+			do_kexec_syscall = 1;
 			break;
 		case OPT_STATUS:
 			do_status = 1;
@@ -1456,16 +1445,16 @@ int main(int argc, char *argv[])
 		result = k_status(kexec_flags);
 	}
 	if (do_unload) {
-		if (do_kexec_file_syscall)
+		if (!do_kexec_syscall)
 			result = kexec_file_unload(kexec_file_flags);
-		else
+		if ((result == -ENOSYS) || !do_kexec_file_syscall)
 			result = k_unload(kexec_flags);
 	}
 	if (do_load && (result == 0)) {
-		if (do_kexec_file_syscall)
+		if (!do_kexec_syscall)
 			result = do_kexec_file_load(fileind, argc, argv,
 						 kexec_file_flags);
-		else
+		if ((result == -ENOSYS) || !do_kexec_file_syscall)
 			result = my_load(type, fileind, argc, argv,
 						kexec_flags, entry);
 	}
kexec/kexec.h +2 -0
index 26225d2c002a..7abcec796cae 100644
@@ -219,6 +219,7 @@ extern int file_types;
 #define OPT_TYPE		't'
 #define OPT_PANIC		'p'
 #define OPT_KEXEC_FILE_SYSCALL	's'
+#define OPT_KEXEC_SYSCALL	'c'
 #define OPT_STATUS		'S'
 #define OPT_MEM_MIN             256
 #define OPT_MEM_MAX             257
@@ -246,6 +247,7 @@ extern int file_types;
 	{ "mem-max",		1, 0, OPT_MEM_MAX }, \
 	{ "reuseinitrd",	0, 0, OPT_REUSE_INITRD }, \
 	{ "kexec-file-syscall",	0, 0, OPT_KEXEC_FILE_SYSCALL }, \
+	{ "kexec-syscall",	0, 0, OPT_KEXEC_SYSCALL }, \
 	{ "debug",		0, 0, OPT_DEBUG }, \
 	{ "status",		0, 0, OPT_STATUS }, \
 	{ "print-ckr-size",     0, 0, OPT_PRINT_CKR_SIZE }, \
Signature · 7 lines
2.13.6


_______________________________________________
kexec mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/kexec    

_______________________________________________
kexec mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/kexec

Quoted 7 lines
Quoted 4 lines

_______________________________________________
kexec mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/kexec

Thanks
Dave

_______________________________________________
kexec mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/kexec

#5 lore raw
Quoted 45 lines · Petr Tesarik
Quoted 42 lines · On Sat, 24 Feb 2018 09:43:42 +0800 Dave Young <[email protected]> wrote:
Quoted 40 lines · Petr Tesarik

Hi Baoquan,

Quoted 6 lines · On Fri, 23 Feb 2018 07:20:43 +0800 Baoquan He <[email protected]> wrote:

Hi Michal,

Quoted 3 lines · Michal Suchanek

The new KEXEC_FILE_LOAD is preferred in the case the platform supports it because it allows kexec in locked down secure boot mode.

Quoted 8 lines
Quoted 2 lines

However, some platforms do not support it so fall back to the old syscall there.

I didn't read code change, just from patch log, I tend to not agree. There are two options KEXEC_FILE_LOAD and KEXEC_LOAD, some platforms do not support, why does some platforms not choose KEXEC_LOAD, the working one? Why bother to make change in code? I believe there's returned message telling if KEXEC_FILE_LOAD works or not.

Well... let me give a bit of background. As you have probably noticed, this syscall was originally available only for x86_64, but more and more architectures are also adding it now.

Next, kexec is actually called by a script (which locates a suitable kernel and initrd, constructs the kernel command line, etc.). The script must either:

A. know somehow if the currently running kernel implements

     kexec_file_load(2), or

B. try one method first, and if it fails, retry with the other.

I agree that kexec(1) should probably allow the user to force a specific method, but I don't see the benefit of implementing fallback in an external script and not in kexec-tools itself.

OTOH if you want to push the fallback logic out of kexec-tools, then I would like to get better diagnostic at least. Letting my script parse kexec output is, um, suboptimal.

Quoted 8 lines
Quoted 2 lines

In Fedora/RHEL we use this in scripts by checking the arch first, for distribution it is enough?

No.

First, you would also have to check the kernel version (and maintain an ugly mapping of which kernel version introduced kexec_file_load on which architecture).

The kernel version update is rare for these new syscall added, but it is indeed needed to match with them

Quoted 8 lines

Second, it's not just the architecture. kexec_load(2) will fail if SecureBoot is active. OTOH kexec_file_load(2) will fail if the kernel is not signed. For kernel hackers who don't use SecureBoot, signing self-built kernels is just overkill. So, you should also check the state of SecureBoot, possibly also whether the kernel image is signed with a valid key, repeating a bit too much of the kernel logic, and quite likely introducing subtle differences...

Hmm, I did not say the exact details, yes, we checked the Secure Boot state and only use kexec_file_load for that special case.

kexec_file and kexec_file_load is not exactly same so if one want to use one instead of another for a specific functionality it seems not good to automatically switch to another if one failed. For example which one should be the first choice, it is hard to say.

Quoted 190 lines

Petr T

Quoted 186 lines

There are also some other arch dependent
options in kexec-tools, there is no way to just use same for every different platform without checking in scripts.

If your scripts is not for a distribution, I agree that it is indeed a problem.

Quoted 173 lines

Petr T

Quoted 159 lines

Thanks
Baoquan

Quoted 157 lines

Also provide an option to call the old syscall in case the new syscall fails with other reason than ENOSYS.

Also document the options.

---
 kexec/kexec.8 |  9 +++++++++
 kexec/kexec.c | 41 +++++++++++++++--------------------------
 kexec/kexec.h |  2 ++
 3 files changed, 26 insertions(+), 26 deletions(-)
kexec/kexec.8 +9 -0
index e0131b4ea827..7e4df723251d 100644
@@ -144,6 +144,15 @@ Load the new kernel for use on panic.
 Specify that the new kernel is of this
 .I type.
 .TP
+.BI \-s\ (\-\-kexec-file-syscall)
+Specify that the new KEXEC_FILE_LOAD syscall should be used exclusively.
+Otherwise KEXEC_FILE_LOAD is tried and when not supported KEXEC_LOAD is used.
+.I type.
+.TP
+.BI \-c\ (\-\-kexec-syscall)
+Specify that the old KEXEC_LOAD syscall should be used exclusively.
+.I type.
+.TP
 .B \-u\ (\-\-unload)
 Unload the current
 .B kexec
kexec/kexec.c +15 -26
index cfd837c1b6bb..25328c02b508 100644
@@ -1166,7 +1166,7 @@ static int do_kexec_file_load(int fileind, int argc, char **argv,
 
 	if (!is_kexec_file_load_implemented()) {
 		fprintf(stderr, "syscall kexec_file_load not available.\n");
-		return -1;
+		return -ENOSYS;
 	}
 
 	if (argc - fileind <= 0) {
@@ -1243,6 +1243,7 @@ int main(int argc, char *argv[])
 	int do_unload = 0;
 	int do_reuse_initrd = 0;
 	int do_kexec_file_syscall = 0;
+	int do_kexec_syscall = 0;
 	int do_status = 0;
 	void *entry = 0;
 	char *type = 0;
@@ -1256,19 +1257,6 @@ int main(int argc, char *argv[])
 	};
 	static const char short_options[] = KEXEC_ALL_OPT_STR;
 
-	/*
-	 * First check if --use-kexec-file-syscall is set. That changes lot of
-	 * things
-	 */
-	while ((opt = getopt_long(argc, argv, short_options,
-				  options, 0)) != -1) {
-		switch(opt) {
-		case OPT_KEXEC_FILE_SYSCALL:
-			do_kexec_file_syscall = 1;
-			break;
-		}
-	}
-
 	/* Reset getopt for the next pass. */
 	opterr = 1;
 	optind = 1;
@@ -1310,8 +1298,7 @@ int main(int argc, char *argv[])
 			do_shutdown = 0;
 			do_sync = 0;
 			do_unload = 1;
-			if (do_kexec_file_syscall)
-				kexec_file_flags |= KEXEC_FILE_UNLOAD;
+			kexec_file_flags |= KEXEC_FILE_UNLOAD;
 			break;
 		case OPT_EXEC:
 			do_load = 0;
@@ -1354,11 +1341,8 @@ int main(int argc, char *argv[])
 			do_exec = 0;
 			do_shutdown = 0;
 			do_sync = 0;
-			if (do_kexec_file_syscall)
-				kexec_file_flags |= KEXEC_FILE_ON_CRASH;
-			else
-				kexec_flags = KEXEC_ON_CRASH;
-			break;
+			kexec_file_flags |= KEXEC_FILE_ON_CRASH;
+			kexec_flags = KEXEC_ON_CRASH;
 		case OPT_MEM_MIN:
 			mem_min = strtoul(optarg, &endptr, 0);
 			if (*endptr) {
@@ -1383,7 +1367,12 @@ int main(int argc, char *argv[])
 			do_reuse_initrd = 1;
 			break;
 		case OPT_KEXEC_FILE_SYSCALL:
-			/* We already parsed it. Nothing to do. */
+			do_kexec_file_syscall = 1;
+			do_kexec_syscall = 0;
+			break;
+		case OPT_KEXEC_SYSCALL:
+			do_kexec_file_syscall = 0;
+			do_kexec_syscall = 1;
 			break;
 		case OPT_STATUS:
 			do_status = 1;
@@ -1456,16 +1445,16 @@ int main(int argc, char *argv[])
 		result = k_status(kexec_flags);
 	}
 	if (do_unload) {
-		if (do_kexec_file_syscall)
+		if (!do_kexec_syscall)
 			result = kexec_file_unload(kexec_file_flags);
-		else
+		if ((result == -ENOSYS) || !do_kexec_file_syscall)
 			result = k_unload(kexec_flags);
 	}
 	if (do_load && (result == 0)) {
-		if (do_kexec_file_syscall)
+		if (!do_kexec_syscall)
 			result = do_kexec_file_load(fileind, argc, argv,
 						 kexec_file_flags);
-		else
+		if ((result == -ENOSYS) || !do_kexec_file_syscall)
 			result = my_load(type, fileind, argc, argv,
 						kexec_flags, entry);
 	}
kexec/kexec.h +2 -0
index 26225d2c002a..7abcec796cae 100644
@@ -219,6 +219,7 @@ extern int file_types;
 #define OPT_TYPE		't'
 #define OPT_PANIC		'p'
 #define OPT_KEXEC_FILE_SYSCALL	's'
+#define OPT_KEXEC_SYSCALL	'c'
 #define OPT_STATUS		'S'
 #define OPT_MEM_MIN             256
 #define OPT_MEM_MAX             257
@@ -246,6 +247,7 @@ extern int file_types;
 	{ "mem-max",		1, 0, OPT_MEM_MAX }, \
 	{ "reuseinitrd",	0, 0, OPT_REUSE_INITRD }, \
 	{ "kexec-file-syscall",	0, 0, OPT_KEXEC_FILE_SYSCALL }, \
+	{ "kexec-syscall",	0, 0, OPT_KEXEC_SYSCALL }, \
 	{ "debug",		0, 0, OPT_DEBUG }, \
 	{ "status",		0, 0, OPT_STATUS }, \
 	{ "print-ckr-size",     0, 0, OPT_PRINT_CKR_SIZE }, \
Signature · 1 line
2.13.6
Quoted 9 lines
Quoted 4 lines

_______________________________________________
kexec mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/kexec

_______________________________________________
kexec mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/kexec

Quoted 4 lines

_______________________________________________
kexec mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/kexec

Thanks
Dave

Thanks
Dave

_______________________________________________
kexec mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/kexec

#6 lore raw
Quoted 70 lines · On Mon, 26 Feb 2018 09:45:15 +0800 Dave Young <[email protected]> wrote:
Quoted 48 lines · Petr Tesarik
Quoted 45 lines · On Sat, 24 Feb 2018 09:43:42 +0800 Dave Young <[email protected]> wrote:
Quoted 43 lines · Petr Tesarik

Hi Baoquan,

Quoted 10 lines · On Fri, 23 Feb 2018 07:20:43 +0800 Baoquan He <[email protected]> wrote:

Hi Michal,

Quoted 6 lines · Michal Suchanek

The new KEXEC_FILE_LOAD is preferred in the case the platform supports it because it allows kexec in locked down secure boot mode.

However, some platforms do not support it so fall back to the old syscall there.

Quoted 6 lines

I didn't read code change, just from patch log, I tend to not agree. There are two options KEXEC_FILE_LOAD and KEXEC_LOAD, some platforms do not support, why does some platforms not choose KEXEC_LOAD, the working one? Why bother to make change in code? I believe there's returned message telling if KEXEC_FILE_LOAD works or not.

Well... let me give a bit of background. As you have probably noticed, this syscall was originally available only for x86_64, but more and more architectures are also adding it now.

Next, kexec is actually called by a script (which locates a suitable kernel and initrd, constructs the kernel command line, etc.). The script must either:

A. know somehow if the currently running kernel implements

     kexec_file_load(2), or

B. try one method first, and if it fails, retry with the
other.

I agree that kexec(1) should probably allow the user to force a specific method, but I don't see the benefit of implementing fallback in an external script and not in kexec-tools itself.

OTOH if you want to push the fallback logic out of kexec-tools, then I would like to get better diagnostic at least. Letting my script parse kexec output is, um, suboptimal.

Quoted 8 lines
Quoted 2 lines

In Fedora/RHEL we use this in scripts by checking the arch first, for distribution it is enough?

No.

First, you would also have to check the kernel version (and maintain an ugly mapping of which kernel version introduced kexec_file_load on which architecture).

The kernel version update is rare for these new syscall added, but it is indeed needed to match with them

Quoted 8 lines

Second, it's not just the architecture. kexec_load(2) will fail if SecureBoot is active. OTOH kexec_file_load(2) will fail if the kernel is not signed. For kernel hackers who don't use SecureBoot, signing self-built kernels is just overkill. So, you should also check the state of SecureBoot, possibly also whether the kernel image is signed with a valid key, repeating a bit too much of the kernel logic, and quite likely introducing subtle differences...

Quoted 7 lines

Hmm, I did not say the exact details, yes, we checked the Secure Boot state and only use kexec_file_load for that special case.

kexec_file and kexec_file_load is not exactly same so if one want to use one instead of another for a specific functionality it seems not good to automatically switch to another if one failed. For example which one should be the first choice, it is hard to say.

Hard to say indeed, However, I would assume that architectures that implement kexec_file_load do so because it is required in some case and hence it will be actively maintained when available. However, some architectures may not require it and will be slow to implement it. So using kexec_file_load when available sounds like the right thing. Technically the implementation details are different but for most users this does not matter.

For those that do care I provide option to select one or the other.

Thanks

Michal

_______________________________________________
kexec mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/kexec

#7 lore raw
Quoted 89 lines · Michal Suchánek
Quoted 70 lines · On Mon, 26 Feb 2018 09:45:15 +0800 Dave Young <[email protected]> wrote:
Quoted 48 lines · Petr Tesarik
Quoted 45 lines · On Sat, 24 Feb 2018 09:43:42 +0800 Dave Young <[email protected]> wrote:
Quoted 43 lines · Petr Tesarik

Hi Baoquan,

Quoted 10 lines · On Fri, 23 Feb 2018 07:20:43 +0800 Baoquan He <[email protected]> wrote:

Hi Michal,

Quoted 6 lines · Michal Suchanek

The new KEXEC_FILE_LOAD is preferred in the case the platform supports it because it allows kexec in locked down secure boot mode.

However, some platforms do not support it so fall back to the old syscall there.

Quoted 6 lines

I didn't read code change, just from patch log, I tend to not agree. There are two options KEXEC_FILE_LOAD and KEXEC_LOAD, some platforms do not support, why does some platforms not choose KEXEC_LOAD, the working one? Why bother to make change in code? I believe there's returned message telling if KEXEC_FILE_LOAD works or not.

Well... let me give a bit of background. As you have probably noticed, this syscall was originally available only for x86_64, but more and more architectures are also adding it now.

Next, kexec is actually called by a script (which locates a suitable kernel and initrd, constructs the kernel command line, etc.). The script must either:

A. know somehow if the currently running kernel implements

     kexec_file_load(2), or

B. try one method first, and if it fails, retry with the
other.

I agree that kexec(1) should probably allow the user to force a specific method, but I don't see the benefit of implementing fallback in an external script and not in kexec-tools itself.

OTOH if you want to push the fallback logic out of kexec-tools, then I would like to get better diagnostic at least. Letting my script parse kexec output is, um, suboptimal.

Quoted 8 lines
Quoted 2 lines

In Fedora/RHEL we use this in scripts by checking the arch first, for distribution it is enough?

No.

First, you would also have to check the kernel version (and maintain an ugly mapping of which kernel version introduced kexec_file_load on which architecture).

The kernel version update is rare for these new syscall added, but it is indeed needed to match with them

Quoted 8 lines

Second, it's not just the architecture. kexec_load(2) will fail if SecureBoot is active. OTOH kexec_file_load(2) will fail if the kernel is not signed. For kernel hackers who don't use SecureBoot, signing self-built kernels is just overkill. So, you should also check the state of SecureBoot, possibly also whether the kernel image is signed with a valid key, repeating a bit too much of the kernel logic, and quite likely introducing subtle differences...

Quoted 7 lines

Hmm, I did not say the exact details, yes, we checked the Secure Boot state and only use kexec_file_load for that special case.

kexec_file and kexec_file_load is not exactly same so if one want to use one instead of another for a specific functionality it seems not good to automatically switch to another if one failed. For example which one should be the first choice, it is hard to say.

Hard to say indeed, However, I would assume that architectures that implement kexec_file_load do so because it is required in some case and hence it will be actively maintained when available. However, some architectures may not require it and will be slow to implement it. So using kexec_file_load when available sounds like the right thing. Technically the implementation details are different but for most users this does not matter.

Quoted 1 line

For those that do care I provide option to select one or the other.

I would say it breaks things, a better way should be introducing another kexec-tools option for example kexec --auto for this purpose. Probably add some --auto=... to select the first chance.

Quoted 9 lines

Thanks

Michal

_______________________________________________
kexec mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/kexec

Thanks
Dave

_______________________________________________
kexec mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/kexec

#8 lore raw
Quoted 90 lines · On Tue, 27 Feb 2018 09:15:10 +0800 Dave Young <[email protected]> wrote:
Quoted 89 lines · Michal Suchánek
Quoted 70 lines · On Mon, 26 Feb 2018 09:45:15 +0800 Dave Young <[email protected]> wrote:
Quoted 48 lines · Petr Tesarik
Quoted 45 lines · On Sat, 24 Feb 2018 09:43:42 +0800 Dave Young <[email protected]> wrote:
Quoted 43 lines · Petr Tesarik

Hi Baoquan,

Quoted 10 lines · On Fri, 23 Feb 2018 07:20:43 +0800 Baoquan He <[email protected]> wrote:

Hi Michal,

Quoted 6 lines · Michal Suchanek

The new KEXEC_FILE_LOAD is preferred in the case the platform supports it because it allows kexec in locked down secure boot mode.

However, some platforms do not support it so fall back to the old syscall there.

Quoted 6 lines

I didn't read code change, just from patch log, I tend to not agree. There are two options KEXEC_FILE_LOAD and KEXEC_LOAD, some platforms do not support, why does some platforms not choose KEXEC_LOAD, the working one? Why bother to make change in code? I believe there's returned message telling if KEXEC_FILE_LOAD works or not.

Well... let me give a bit of background. As you have probably noticed, this syscall was originally available only for x86_64, but more and more architectures are also adding it now.

Next, kexec is actually called by a script (which locates a suitable kernel and initrd, constructs the kernel command line, etc.). The script must either:

A. know somehow if the currently running kernel implements

     kexec_file_load(2), or

B. try one method first, and if it fails, retry with the
other.

I agree that kexec(1) should probably allow the user to force a specific method, but I don't see the benefit of implementing fallback in an external script and not in kexec-tools itself.

OTOH if you want to push the fallback logic out of kexec-tools, then I would like to get better diagnostic at least. Letting my script parse kexec output is, um, suboptimal.

Quoted 8 lines
Quoted 2 lines

In Fedora/RHEL we use this in scripts by checking the arch first, for distribution it is enough?

No.

First, you would also have to check the kernel version (and maintain an ugly mapping of which kernel version introduced kexec_file_load on which architecture).

The kernel version update is rare for these new syscall added, but it is indeed needed to match with them

Quoted 8 lines

Second, it's not just the architecture. kexec_load(2) will fail if SecureBoot is active. OTOH kexec_file_load(2) will fail if the kernel is not signed. For kernel hackers who don't use SecureBoot, signing self-built kernels is just overkill. So, you should also check the state of SecureBoot, possibly also whether the kernel image is signed with a valid key, repeating a bit too much of the kernel logic, and quite likely introducing subtle differences...

Quoted 7 lines

Hmm, I did not say the exact details, yes, we checked the Secure Boot state and only use kexec_file_load for that special case.

kexec_file and kexec_file_load is not exactly same so if one want to use one instead of another for a specific functionality it seems not good to automatically switch to another if one failed. For example which one should be the first choice, it is hard to say.

Hard to say indeed, However, I would assume that architectures that implement kexec_file_load do so because it is required in some case and hence it will be actively maintained when available. However, some architectures may not require it and will be slow to implement it. So using kexec_file_load when available sounds like the right thing. Technically the implementation details are different but for most users this does not matter.

Quoted 5 lines
Quoted 1 line

For those that do care I provide option to select one or the other.

I would say it breaks things, a better way should be introducing another kexec-tools option for example kexec --auto for this purpose. Probably add some --auto=... to select the first chance.

Yes! This is the way to go. But then I wouldn't call it --auto. I would call it "--method=", so you could specify:

--method=kernel (to use the in-kernel loader aka kexec_file_load)
--method=user (to use the traditional user-space loader)
--method=kernel,user (to prefer kexec_file_load) --method=user,kernel (to prefer kexc_load)

I'm not quite sure if it also makes sense to provide "--method=auto", which would use whatever default is considered sane for the running system.

Petr T

_______________________________________________
kexec mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/kexec

#9 lore raw
Quoted 102 lines · Petr Tesarik
Quoted 90 lines · On Tue, 27 Feb 2018 09:15:10 +0800 Dave Young <[email protected]> wrote:
Quoted 89 lines · Michal Suchánek
Quoted 70 lines · On Mon, 26 Feb 2018 09:45:15 +0800 Dave Young <[email protected]> wrote:
Quoted 48 lines · Petr Tesarik
Quoted 45 lines · On Sat, 24 Feb 2018 09:43:42 +0800 Dave Young <[email protected]> wrote:
Quoted 43 lines · Petr Tesarik

Hi Baoquan,

Quoted 10 lines · On Fri, 23 Feb 2018 07:20:43 +0800 Baoquan He <[email protected]> wrote:

Hi Michal,

Quoted 6 lines · Michal Suchanek

The new KEXEC_FILE_LOAD is preferred in the case the platform supports it because it allows kexec in locked down secure boot mode.

However, some platforms do not support it so fall back to the old syscall there.

Quoted 6 lines

I didn't read code change, just from patch log, I tend to not agree. There are two options KEXEC_FILE_LOAD and KEXEC_LOAD, some platforms do not support, why does some platforms not choose KEXEC_LOAD, the working one? Why bother to make change in code? I believe there's returned message telling if KEXEC_FILE_LOAD works or not.

Well... let me give a bit of background. As you have probably noticed, this syscall was originally available only for x86_64, but more and more architectures are also adding it now.

Next, kexec is actually called by a script (which locates a suitable kernel and initrd, constructs the kernel command line, etc.). The script must either:

A. know somehow if the currently running kernel implements

     kexec_file_load(2), or

B. try one method first, and if it fails, retry with the
other.

I agree that kexec(1) should probably allow the user to force a specific method, but I don't see the benefit of implementing fallback in an external script and not in kexec-tools itself.

OTOH if you want to push the fallback logic out of kexec-tools, then I would like to get better diagnostic at least. Letting my script parse kexec output is, um, suboptimal.

Quoted 8 lines
Quoted 2 lines

In Fedora/RHEL we use this in scripts by checking the arch first, for distribution it is enough?

No.

First, you would also have to check the kernel version (and maintain an ugly mapping of which kernel version introduced kexec_file_load on which architecture).

The kernel version update is rare for these new syscall added, but it is indeed needed to match with them

Quoted 8 lines

Second, it's not just the architecture. kexec_load(2) will fail if SecureBoot is active. OTOH kexec_file_load(2) will fail if the kernel is not signed. For kernel hackers who don't use SecureBoot, signing self-built kernels is just overkill. So, you should also check the state of SecureBoot, possibly also whether the kernel image is signed with a valid key, repeating a bit too much of the kernel logic, and quite likely introducing subtle differences...

Quoted 7 lines

Hmm, I did not say the exact details, yes, we checked the Secure Boot state and only use kexec_file_load for that special case.

kexec_file and kexec_file_load is not exactly same so if one want to use one instead of another for a specific functionality it seems not good to automatically switch to another if one failed. For example which one should be the first choice, it is hard to say.

Hard to say indeed, However, I would assume that architectures that implement kexec_file_load do so because it is required in some case and hence it will be actively maintained when available. However, some architectures may not require it and will be slow to implement it. So using kexec_file_load when available sounds like the right thing. Technically the implementation details are different but for most users this does not matter.

Quoted 5 lines
Quoted 1 line

For those that do care I provide option to select one or the other.

I would say it breaks things, a better way should be introducing another kexec-tools option for example kexec --auto for this purpose. Probably add some --auto=... to select the first chance.

Yes! This is the way to go. But then I wouldn't call it --auto. I would call it "--method=", so you could specify:

Quoted 8 lines

--method=kernel (to use the in-kernel loader aka kexec_file_load)
--method=user (to use the traditional user-space loader)
--method=kernel,user (to prefer kexec_file_load) --method=user,kernel (to prefer kexc_load)

I'm not quite sure if it also makes sense to provide "--method=auto", which would use whatever default is considered sane for the running system.

If we have --method=kernel,user and --method=user,kernel then no need "auto" anymore.. But I would prefer that original default behavior is kept, kexec -l without "--method" will use userspace loader, kexec -s -l will use kernel space loader

Quoted 2 lines

Petr T

Thanks
Dave

_______________________________________________
kexec mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/kexec

#10 lore raw
Quoted 108 lines · Dave Young
Quoted 107 lines · Petr Tesarik
Quoted 90 lines · On Tue, 27 Feb 2018 09:15:10 +0800 Dave Young <[email protected]> wrote:
Quoted 89 lines · Michal Suchánek
Quoted 70 lines · On Mon, 26 Feb 2018 09:45:15 +0800 Dave Young <[email protected]> wrote:
Quoted 48 lines · Petr Tesarik
Quoted 45 lines · On Sat, 24 Feb 2018 09:43:42 +0800 Dave Young <[email protected]> wrote:
Quoted 43 lines · Petr Tesarik

Hi Baoquan,

Quoted 10 lines · On Fri, 23 Feb 2018 07:20:43 +0800 Baoquan He <[email protected]> wrote:

Hi Michal,

Quoted 6 lines · Michal Suchanek

The new KEXEC_FILE_LOAD is preferred in the case the platform supports it because it allows kexec in locked down secure boot mode.

However, some platforms do not support it so fall back to the old syscall there.

Quoted 6 lines

I didn't read code change, just from patch log, I tend to not agree. There are two options KEXEC_FILE_LOAD and KEXEC_LOAD, some platforms do not support, why does some platforms not choose KEXEC_LOAD, the working one? Why bother to make change in code? I believe there's returned message telling if KEXEC_FILE_LOAD works or not.

Well... let me give a bit of background. As you have probably noticed, this syscall was originally available only for x86_64, but more and more architectures are also adding it now.

Next, kexec is actually called by a script (which locates a suitable kernel and initrd, constructs the kernel command line, etc.). The script must either:

A. know somehow if the currently running kernel implements

     kexec_file_load(2), or

B. try one method first, and if it fails, retry with the
other.

I agree that kexec(1) should probably allow the user to force a specific method, but I don't see the benefit of implementing fallback in an external script and not in kexec-tools itself.

OTOH if you want to push the fallback logic out of kexec-tools, then I would like to get better diagnostic at least. Letting my script parse kexec output is, um, suboptimal.

Quoted 8 lines
Quoted 2 lines

In Fedora/RHEL we use this in scripts by checking the arch first, for distribution it is enough?

No.

First, you would also have to check the kernel version (and maintain an ugly mapping of which kernel version introduced kexec_file_load on which architecture).

The kernel version update is rare for these new syscall added, but it is indeed needed to match with them

Quoted 8 lines

Second, it's not just the architecture. kexec_load(2) will fail if SecureBoot is active. OTOH kexec_file_load(2) will fail if the kernel is not signed. For kernel hackers who don't use SecureBoot, signing self-built kernels is just overkill. So, you should also check the state of SecureBoot, possibly also whether the kernel image is signed with a valid key, repeating a bit too much of the kernel logic, and quite likely introducing subtle differences...

Quoted 7 lines

Hmm, I did not say the exact details, yes, we checked the Secure Boot state and only use kexec_file_load for that special case.

kexec_file and kexec_file_load is not exactly same so if one want to use one instead of another for a specific functionality it seems not good to automatically switch to another if one failed. For example which one should be the first choice, it is hard to say.

Hard to say indeed, However, I would assume that architectures that implement kexec_file_load do so because it is required in some case and hence it will be actively maintained when available. However, some architectures may not require it and will be slow to implement it. So using kexec_file_load when available sounds like the right thing. Technically the implementation details are different but for most users this does not matter.

Quoted 5 lines
Quoted 1 line

For those that do care I provide option to select one or the other.

I would say it breaks things, a better way should be introducing another kexec-tools option for example kexec --auto for this purpose. Probably add some --auto=... to select the first chance.

Yes! This is the way to go. But then I wouldn't call it --auto. I would call it "--method=", so you could specify:

--method=kernel (to use the in-kernel loader aka kexec_file_load)
--method=user (to use the traditional user-space loader)
--method=kernel,user (to prefer kexec_file_load) --method=user,kernel (to prefer kexc_load)

Quoted 8 lines
Quoted 3 lines

I'm not quite sure if it also makes sense to provide "--method=auto", which would use whatever default is considered sane for the running system.

If we have --method=kernel,user and --method=user,kernel then no need "auto" anymore.. But I would prefer that original default behavior is kept, kexec -l without "--method" will use userspace loader, kexec -s -l will use kernel space loader

If one use kexec -s --method=user then kexec-tools should fail, talked with Baoquan privately he has another idea to extend the kexec -s with --fallback: "kexec -s --fallback -l" can fallback to userspace loader this seems also a good option.

Quoted 6 lines
Quoted 3 lines

Petr T

Thanks
Dave

_______________________________________________
kexec mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/kexec

#11 lore raw
Quoted 130 lines · On Tue, 27 Feb 2018 17:14:02 +0800 Dave Young <[email protected]> wrote:
Quoted 120 lines · Dave Young
Quoted 119 lines · Petr Tesarik
Quoted 100 lines · On Tue, 27 Feb 2018 09:15:10 +0800 Dave Young <[email protected]> wrote:
Quoted 99 lines · Michal Suchánek
Quoted 81 lines · On Mon, 26 Feb 2018 09:45:15 +0800 Dave Young <[email protected]> wrote:
Quoted 53 lines · Petr Tesarik
Quoted 50 lines · On Sat, 24 Feb 2018 09:43:42 +0800 Dave Young <[email protected]> wrote:
Quoted 48 lines · Petr Tesarik

Hi Baoquan,

Quoted 11 lines · On Fri, 23 Feb 2018 07:20:43 +0800 Baoquan He <[email protected]> wrote:

Hi Michal,

Quoted 6 lines · Michal Suchanek

The new KEXEC_FILE_LOAD is preferred in the case the platform supports it because it allows kexec in locked down secure boot mode.

However, some platforms do not support it so fall back to the old syscall there.

Quoted 7 lines

I didn't read code change, just from patch log, I tend to not agree. There are two options KEXEC_FILE_LOAD and KEXEC_LOAD, some platforms do not support, why does some platforms not choose KEXEC_LOAD, the working one? Why bother to make change in code? I believe there's returned message telling if KEXEC_FILE_LOAD works or not.

Well... let me give a bit of background. As you have probably noticed, this syscall was originally available only for x86_64, but more and more architectures are also adding it now.

Next, kexec is actually called by a script (which locates a suitable kernel and initrd, constructs the kernel command line, etc.). The script must either:

A. know somehow if the currently running kernel
implements kexec_file_load(2), or

B. try one method first, and if it fails, retry
with the other.

I agree that kexec(1) should probably allow the user to force a specific method, but I don't see the benefit of implementing fallback in an external script and not in kexec-tools itself.

OTOH if you want to push the fallback logic out of kexec-tools, then I would like to get better diagnostic at least. Letting my script parse kexec output is, um, suboptimal.

Quoted 8 lines
Quoted 2 lines

In Fedora/RHEL we use this in scripts by checking the arch first, for distribution it is enough?

No.

First, you would also have to check the kernel version (and maintain an ugly mapping of which kernel version introduced kexec_file_load on which architecture).

The kernel version update is rare for these new syscall added, but it is indeed needed to match with them

Quoted 10 lines

Second, it's not just the architecture. kexec_load(2) will fail if SecureBoot is active. OTOH kexec_file_load(2) will fail if the kernel is not signed. For kernel hackers who don't use SecureBoot, signing self-built kernels is just overkill. So, you should also check the state of SecureBoot, possibly also whether the kernel image is signed with a valid key, repeating a bit too much of the kernel logic, and quite likely introducing subtle differences...

Hmm, I did not say the exact details, yes, we checked the Secure Boot state and only use kexec_file_load for that special case.

Quoted 5 lines

kexec_file and kexec_file_load is not exactly same so if one want to use one instead of another for a specific functionality it seems not good to automatically switch to another if one failed. For example which one should be the first choice, it is hard to say.

Hard to say indeed, However, I would assume that architectures that implement kexec_file_load do so because it is required in some case and hence it will be actively maintained when available. However, some architectures may not require it and will be slow to implement it. So using kexec_file_load when available sounds like the right thing. Technically the implementation details are different but for most users this does not matter.

Quoted 7 lines
Quoted 2 lines

For those that do care I provide option to select one or the other.

I would say it breaks things, a better way should be introducing another kexec-tools option for example kexec --auto for this purpose. Probably add some --auto=... to select the first chance.

Yes! This is the way to go. But then I wouldn't call it --auto. I would call it "--method=", so you could specify:

--method=kernel (to use the in-kernel loader aka
kexec_file_load) --method=user (to use the traditional
user-space loader) --method=kernel,user (to prefer kexec_file_load) --method=user,kernel (to prefer kexc_load)

Quoted 8 lines
Quoted 3 lines

I'm not quite sure if it also makes sense to provide "--method=auto", which would use whatever default is considered sane for the running system.

If we have --method=kernel,user and --method=user,kernel then no need "auto" anymore.. But I would prefer that original default behavior is kept, kexec -l without "--method" will use userspace loader, kexec -s -l will use kernel space loader

Quoted 4 lines

If one use kexec -s --method=user then kexec-tools should fail, talked with Baoquan privately he has another idea to extend the kexec -s with --fallback: "kexec -s --fallback -l" can fallback to userspace loader this seems also a good option.

In the updated series it is easy to turn on or off the fallback so an option can be added for that. However, the fallback only works by trying -s first. Implementing fallback in either direction would complicate things a bit. Especially since error codes of kexec-tools do not tend to be nice and consistent.

Thanks

Michal

_______________________________________________
kexec mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/kexec

#12 lore raw
In reply to Dave Young

Hello,

Quoted 46 lines · On Sat, 24 Feb 2018 10:19:09 +0800 Baoquan He <[email protected]> wrote:

Hi Petr,

Quoted 35 lines · Petr Tesarik

Hi Baoquan,

Quoted 10 lines · On Fri, 23 Feb 2018 07:20:43 +0800 Baoquan He <[email protected]> wrote:

Hi Michal,

Quoted 6 lines · Michal Suchanek

The new KEXEC_FILE_LOAD is preferred in the case the platform supports it because it allows kexec in locked down secure boot mode.

However, some platforms do not support it so fall back to the old syscall there.

Quoted 6 lines

I didn't read code change, just from patch log, I tend to not agree. There are two options KEXEC_FILE_LOAD and KEXEC_LOAD, some platforms do not support, why does some platforms not choose KEXEC_LOAD, the working one? Why bother to make change in code? I believe there's returned message telling if KEXEC_FILE_LOAD works or not.

Well... let me give a bit of background. As you have probably noticed, this syscall was originally available only for x86_64, but more and more architectures are also adding it now.

Next, kexec is actually called by a script (which locates a suitable kernel and initrd, constructs the kernel command line, etc.). The script must either:

A. know somehow if the currently running kernel implements

     kexec_file_load(2), or

B. try one method first, and if it fails, retry with the other.

Quoted 7 lines

I agree that kexec(1) should probably allow the user to force a specific method, but I don't see the benefit of implementing fallback in an external script and not in kexec-tools itself.

OTOH if you want to push the fallback logic out of kexec-tools, then I would like to get better diagnostic at least. Letting my script parse kexec output is, um, suboptimal.

Quoted 5 lines

Thanks for the details!

Firstly, this patch is very ugly. It mixs the falling back issue, doc adding, code cleanup in one patch. It's not easy to see how many lines of code change involved.

The -s option is special-cased and I needed to un-special-case it to be able to use either syscall. I suppose this can be split.

Quoted 2 lines

Secondly, I personally like better the A or B two options.

Which the patch does provide. If you want A you can select it. if you want B you can select it as well. If you do not select either kexec tries A and if not supported tries B.

Quoted 2 lines

For A,
checking ARCH in script might be a easier thing.

What does checking ARCH in a script tell you? You need to check that the syscall is supported both in kexec and the running kernel. And only kexec can do that.

Quoted 3 lines

And for B, just check
if "syscall kexec_file_load not available" is printed, then retry the KEXEC_LOAD.

Oh right, that's totally robust programming. If kexec does not know how to call KEXEC_FILE_LOAD then the message is printed to stderr and -1 is returned whereas when kexec thinks it knows how to call KEXEC_FILE_LOAD and the kernel does not support it nothing is printed and the return code from kernel is passed through which is hopefully -ENOSYS.

Can we at least make the return code consistent?

Quoted 3 lines

And the falling back may give people a feeling that that ARCH support the file mode loading.

Why? How?

Previously they had to select KEXEC_FILE_LOAD with an undocumented option. It is now documented and when used the fallback is disabled - no change.

Quoted 5 lines

Besides, if fall back to KEXEC_LOAD when
KEXEC_FILE_LOAD is tried but not supported this time, next time people may want to do fall back when KEXEC_FILE_LOAD is tried but failed, since there has been a precedent. This might be not good for keeping code logic simple, add complexity to maintaining.

What complexity are you talking about? That is exactly what the fallback does except it checks that the error code was ENOSYS. It might make sense to also use the fallback when the file format is not recognized by the kernel. KEXEC_FILE_LOAD currently does not support multiboot and uImage formats. So if a specific error code is returned in this case (as opposed to known image type which is invalid) it might make sense to use the fallback as well.

Quoted 6 lines

My personal opinion.

Thanks
Baoquan

Quoted 40 lines · On Sat, 24 Feb 2018 09:43:42 +0800 Dave Young <[email protected]> wrote:
Quoted 39 lines · Petr Tesarik

Hi Baoquan,

Quoted 10 lines · On Fri, 23 Feb 2018 07:20:43 +0800 Baoquan He <[email protected]> wrote:

Hi Michal,

Quoted 6 lines · Michal Suchanek

The new KEXEC_FILE_LOAD is preferred in the case the platform supports it because it allows kexec in locked down secure boot mode.

However, some platforms do not support it so fall back to the old syscall there.

Quoted 6 lines

I didn't read code change, just from patch log, I tend to not agree. There are two options KEXEC_FILE_LOAD and KEXEC_LOAD, some platforms do not support, why does some platforms not choose KEXEC_LOAD, the working one? Why bother to make change in code? I believe there's returned message telling if KEXEC_FILE_LOAD works or not.

Well... let me give a bit of background. As you have probably noticed, this syscall was originally available only for x86_64, but more and more architectures are also adding it now.

Next, kexec is actually called by a script (which locates a suitable kernel and initrd, constructs the kernel command line, etc.). The script must either:

A. know somehow if the currently running kernel implements

     kexec_file_load(2), or

B. try one method first, and if it fails, retry with the other.

I agree that kexec(1) should probably allow the user to force a specific method, but I don't see the benefit of implementing fallback in an external script and not in kexec-tools itself.

Quoted 6 lines
Quoted 3 lines

OTOH if you want to push the fallback logic out of kexec-tools, then I would like to get better diagnostic at least. Letting my script parse kexec output is, um, suboptimal.

In Fedora/RHEL we use this in scripts by checking the arch first, for distribution it is enough?

No. That means you need to update the script when you implement KEXEC_FILE_LOAD on a new arch. Worse, you need to check the kernel version so it works on old kernel as well.

Quoted 3 lines

There are also some other arch
dependent options in kexec-tools, there is no way to just use same for every different platform without checking in scripts.

Like?

And here I mean options that are specific to the platform and not a particular system. Because system-specific options have to be set case by case, obviously. I see no kexec option that would generally apply on one arch and not other except -s.

Thanks

Michal

_______________________________________________
kexec mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/kexec

#13 lore raw
In reply to Petr Tesarik

Hi Petr,

Quoted 33 lines · Petr Tesarik

Hi Baoquan,

Quoted 6 lines · On Fri, 23 Feb 2018 07:20:43 +0800 Baoquan He <[email protected]> wrote:

Hi Michal,

Quoted 3 lines · Michal Suchanek

The new KEXEC_FILE_LOAD is preferred in the case the platform supports it because it allows kexec in locked down secure boot mode.

Quoted 8 lines
Quoted 2 lines

However, some platforms do not support it so fall back to the old syscall there.

I didn't read code change, just from patch log, I tend to not agree. There are two options KEXEC_FILE_LOAD and KEXEC_LOAD, some platforms do not support, why does some platforms not choose KEXEC_LOAD, the working one? Why bother to make change in code? I believe there's returned message telling if KEXEC_FILE_LOAD works or not.

Well... let me give a bit of background. As you have probably noticed, this syscall was originally available only for x86_64, but more and more architectures are also adding it now.

Next, kexec is actually called by a script (which locates a suitable kernel and initrd, constructs the kernel command line, etc.). The script must either:

A. know somehow if the currently running kernel implements

     kexec_file_load(2), or

B. try one method first, and if it fails, retry with the other.

Quoted 7 lines

I agree that kexec(1) should probably allow the user to force a specific method, but I don't see the benefit of implementing fallback in an external script and not in kexec-tools itself.

OTOH if you want to push the fallback logic out of kexec-tools, then I would like to get better diagnostic at least. Letting my script parse kexec output is, um, suboptimal.

Thanks for the details!

Firstly, this patch is very ugly. It mixs the falling back issue, doc adding, code cleanup in one patch. It's not easy to see how many lines of code change involved.

Secondly, I personally like better the A or B two options. For A, checking ARCH in script might be a easier thing. And for B, just check if "syscall kexec_file_load not available" is printed, then retry the KEXEC_LOAD.

And the falling back may give people a feeling that that ARCH support the file mode loading. Besides, if fall back to KEXEC_LOAD when KEXEC_FILE_LOAD is tried but not supported this time, next time people may want to do fall back when KEXEC_FILE_LOAD is tried but failed, since there has been a precedent. This might be not good for keeping code logic simple, add complexity to maintaining.

My personal opinion.

Thanks
Baoquan

Quoted 167 lines
Quoted 156 lines
Quoted 156 lines

Also provide an option to call the old syscall in case the new syscall fails with other reason than ENOSYS.

Also document the options.

---
 kexec/kexec.8 |  9 +++++++++
 kexec/kexec.c | 41 +++++++++++++++--------------------------
 kexec/kexec.h |  2 ++
 3 files changed, 26 insertions(+), 26 deletions(-)
kexec/kexec.8 +9 -0
index e0131b4ea827..7e4df723251d 100644
@@ -144,6 +144,15 @@ Load the new kernel for use on panic.
 Specify that the new kernel is of this
 .I type.
 .TP
+.BI \-s\ (\-\-kexec-file-syscall)
+Specify that the new KEXEC_FILE_LOAD syscall should be used exclusively.
+Otherwise KEXEC_FILE_LOAD is tried and when not supported KEXEC_LOAD is used.
+.I type.
+.TP
+.BI \-c\ (\-\-kexec-syscall)
+Specify that the old KEXEC_LOAD syscall should be used exclusively.
+.I type.
+.TP
 .B \-u\ (\-\-unload)
 Unload the current
 .B kexec
kexec/kexec.c +15 -26
index cfd837c1b6bb..25328c02b508 100644
@@ -1166,7 +1166,7 @@ static int do_kexec_file_load(int fileind, int argc, char **argv,
 
 	if (!is_kexec_file_load_implemented()) {
 		fprintf(stderr, "syscall kexec_file_load not available.\n");
-		return -1;
+		return -ENOSYS;
 	}
 
 	if (argc - fileind <= 0) {
@@ -1243,6 +1243,7 @@ int main(int argc, char *argv[])
 	int do_unload = 0;
 	int do_reuse_initrd = 0;
 	int do_kexec_file_syscall = 0;
+	int do_kexec_syscall = 0;
 	int do_status = 0;
 	void *entry = 0;
 	char *type = 0;
@@ -1256,19 +1257,6 @@ int main(int argc, char *argv[])
 	};
 	static const char short_options[] = KEXEC_ALL_OPT_STR;
 
-	/*
-	 * First check if --use-kexec-file-syscall is set. That changes lot of
-	 * things
-	 */
-	while ((opt = getopt_long(argc, argv, short_options,
-				  options, 0)) != -1) {
-		switch(opt) {
-		case OPT_KEXEC_FILE_SYSCALL:
-			do_kexec_file_syscall = 1;
-			break;
-		}
-	}
-
 	/* Reset getopt for the next pass. */
 	opterr = 1;
 	optind = 1;
@@ -1310,8 +1298,7 @@ int main(int argc, char *argv[])
 			do_shutdown = 0;
 			do_sync = 0;
 			do_unload = 1;
-			if (do_kexec_file_syscall)
-				kexec_file_flags |= KEXEC_FILE_UNLOAD;
+			kexec_file_flags |= KEXEC_FILE_UNLOAD;
 			break;
 		case OPT_EXEC:
 			do_load = 0;
@@ -1354,11 +1341,8 @@ int main(int argc, char *argv[])
 			do_exec = 0;
 			do_shutdown = 0;
 			do_sync = 0;
-			if (do_kexec_file_syscall)
-				kexec_file_flags |= KEXEC_FILE_ON_CRASH;
-			else
-				kexec_flags = KEXEC_ON_CRASH;
-			break;
+			kexec_file_flags |= KEXEC_FILE_ON_CRASH;
+			kexec_flags = KEXEC_ON_CRASH;
 		case OPT_MEM_MIN:
 			mem_min = strtoul(optarg, &endptr, 0);
 			if (*endptr) {
@@ -1383,7 +1367,12 @@ int main(int argc, char *argv[])
 			do_reuse_initrd = 1;
 			break;
 		case OPT_KEXEC_FILE_SYSCALL:
-			/* We already parsed it. Nothing to do. */
+			do_kexec_file_syscall = 1;
+			do_kexec_syscall = 0;
+			break;
+		case OPT_KEXEC_SYSCALL:
+			do_kexec_file_syscall = 0;
+			do_kexec_syscall = 1;
 			break;
 		case OPT_STATUS:
 			do_status = 1;
@@ -1456,16 +1445,16 @@ int main(int argc, char *argv[])
 		result = k_status(kexec_flags);
 	}
 	if (do_unload) {
-		if (do_kexec_file_syscall)
+		if (!do_kexec_syscall)
 			result = kexec_file_unload(kexec_file_flags);
-		else
+		if ((result == -ENOSYS) || !do_kexec_file_syscall)
 			result = k_unload(kexec_flags);
 	}
 	if (do_load && (result == 0)) {
-		if (do_kexec_file_syscall)
+		if (!do_kexec_syscall)
 			result = do_kexec_file_load(fileind, argc, argv,
 						 kexec_file_flags);
-		else
+		if ((result == -ENOSYS) || !do_kexec_file_syscall)
 			result = my_load(type, fileind, argc, argv,
 						kexec_flags, entry);
 	}
kexec/kexec.h +2 -0
index 26225d2c002a..7abcec796cae 100644
@@ -219,6 +219,7 @@ extern int file_types;
 #define OPT_TYPE		't'
 #define OPT_PANIC		'p'
 #define OPT_KEXEC_FILE_SYSCALL	's'
+#define OPT_KEXEC_SYSCALL	'c'
 #define OPT_STATUS		'S'
 #define OPT_MEM_MIN             256
 #define OPT_MEM_MAX             257
@@ -246,6 +247,7 @@ extern int file_types;
 	{ "mem-max",		1, 0, OPT_MEM_MAX }, \
 	{ "reuseinitrd",	0, 0, OPT_REUSE_INITRD }, \
 	{ "kexec-file-syscall",	0, 0, OPT_KEXEC_FILE_SYSCALL }, \
+	{ "kexec-syscall",	0, 0, OPT_KEXEC_SYSCALL }, \
 	{ "debug",		0, 0, OPT_DEBUG }, \
 	{ "status",		0, 0, OPT_STATUS }, \
 	{ "print-ckr-size",     0, 0, OPT_PRINT_CKR_SIZE }, \
Signature · 1 line
2.13.6
Quoted 9 lines
Quoted 4 lines

_______________________________________________
kexec mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/kexec

_______________________________________________
kexec mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/kexec

Quoted 4 lines

_______________________________________________
kexec mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/kexec

_______________________________________________
kexec mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/kexec

#14 lore raw
In reply to Baoquan He
Quoted 6 lines · On Fri, 23 Feb 2018 07:20:43 +0800 Baoquan He <[email protected]> wrote:

Hi Michal,

Quoted 3 lines · Michal Suchanek

The new KEXEC_FILE_LOAD is preferred in the case the platform supports it because it allows kexec in locked down secure boot mode.

Quoted 7 lines
Quoted 2 lines

However, some platforms do not support it so fall back to the old syscall there.

I didn't read code change, just from patch log, I tend to not agree. There are two options KEXEC_FILE_LOAD and KEXEC_LOAD, some platforms do not support, why does some platforms not choose KEXEC_LOAD, the working one?

Because nobody wrote the support. If you volunteer to write support for KEXEC_FILE_LOAD for every platform Linux supports and add the respective syscall numbers to kexec so it knows how to execute the syscall on every platform I will consider it alternative fix.

Some people will argue that not everyone applies the patches to support KEXEC_FILE_LOAD in the kernel overnight, though.

Quoted 1 line

Why bother to make change in code?

Because it is unusable as is. Just calling kexec fails with locked-down secure boot. Calling kexec -s fails on almost every platform except x86.

Quoted 2 lines

I believe there's
returned message telling if KEXEC_FILE_LOAD works or not.

That is not a solution. A way to call kexec that actually works is needed. This patch removes the need to use the undocumented -s option to get the new superior syscall you seem to prefer. It will just do the right thing in most cases. It allows the user to select either syscall explicitly as well. I do not see the problem with that.

Thanks

Michal

_______________________________________________
kexec mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/kexec