Skip to contentSkip to conversation

libswap.c: Improve caculate swap dev number

PATCH Superseded Wei Gao via ltp 3 participants 14 messages Mar 1, 2024 – Mar 6, 2024

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 libswap.c: Improve caculate swap dev number

I encounter a dead loop on following code in our test on ppc64 machine:
while ((c = fgetc(fp)) != EOF)

I think "/proc/swaps" is not normal file and can not get EOF in some situation, so i use fgets a line and caculate swap dev number.

---
 libs/libltpswap/libswap.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)
libs/libltpswap/libswap.c +6 -4
index c8cbb8ea1..6924066b7 100644
@@ -13,6 +13,7 @@
 
 #define TST_NO_DEFAULT_MAIN
 #define DEFAULT_MAX_SWAPFILE 32
+#define MAX_LINE_LEN 256
 
 #include "tst_test.h"
 #include "libswap.h"
@@ -274,16 +275,17 @@ int tst_max_swapfiles(void)
 int tst_count_swaps(void)
 {
 	FILE *fp;
-	int used = -1;
-	char c;
+	int used = 0;
 
 	fp = SAFE_FOPEN("/proc/swaps", "r");
 	if (fp == NULL)
 		return -1;
 
-	while ((c = fgetc(fp)) != EOF) {
-		if (c == '\n')
+	char line[MAX_LINE_LEN];
+	while (fgets(line, MAX_LINE_LEN, fp) != NULL) {
+		if (strstr(line, "/dev/") != NULL) {
 			used++;
+		}
 	}
 
 	SAFE_FCLOSE(fp);
Signature · 5 lines
2.35.3


-- 
Mailing list info: https://lists.linux.it/listinfo/ltp
#1 lore raw

Hi Wei

Quoted 33 lines

I encounter a dead loop on following code in our test on ppc64 machine:
while ((c = fgetc(fp)) != EOF)

I think "/proc/swaps" is not normal file and can not get EOF in some situation, so i use fgets a line and caculate swap dev number.

---
  libs/libltpswap/libswap.c | 10 ++++++----
  1 file changed, 6 insertions(+), 4 deletions(-)
libs/libltpswap/libswap.c +2 -2
index c8cbb8ea1..6924066b7 100644
@@ -13,6 +13,7 @@
 
  #define TST_NO_DEFAULT_MAIN
  #define DEFAULT_MAX_SWAPFILE 32
+#define MAX_LINE_LEN 256
 
  #include "tst_test.h"
  #include "libswap.h"
@@ -274,16 +275,17 @@ int tst_max_swapfiles(void)
  int tst_count_swaps(void)
  {
  	FILE *fp;
-	int used = -1;
-	char c;
+	int used = 0;
 
  	fp = SAFE_FOPEN("/proc/swaps", "r");
  	if (fp == NULL)

libs/libltpswap/libswap.c int tst_max_swapfiles(void) · line 289

Quoted 10 lines
libs/libltpswap/libswap.c +4 -2
  		return -1;
 
-	while ((c = fgetc(fp)) != EOF) {
-		if (c == '\n')
+	char line[MAX_LINE_LEN];
+	while (fgets(line, MAX_LINE_LEN, fp) != NULL) {
+		if (strstr(line, "/dev/") != NULL) {
  			used++;
+		}
  	}

You are not the first person to meet this deadloop problem, Petr also met this[1] in my v4 patch..

But I don't think it related to /proc/swapfiles, I doubot libc wrapper for fgetc problem on ppc64 machine.

Can you try fgetc problem by using fgetc api in ipc library[2]? Then we can know the right reason whether is /proc/swaps or getc problem.

If so, I think we can change this as my v2 way[3].

[1]https://patchwork.ozlabs.org/project/ltp/patch/[email protected]/

[2]https://github.com/linux-test-project/ltp/blob/master/libs/libltpnewipc/libnewipc.c#L58

[3]https://patchwork.ozlabs.org/project/ltp/patch/[email protected]/

Best Regards
Yang Xu

Quoted 2 lines
libs/libltpswap/libswap.c +0 -0
  	SAFE_FCLOSE(fp);
Signature · 1 line
Mailing list info: https://lists.linux.it/listinfo/ltp
#2 lore raw

Hi Wei, Xu,

Quoted 7 lines

Hi Wei

Quoted 5 lines

I encounter a dead loop on following code in our test on ppc64 machine:
while ((c = fgetc(fp)) != EOF)

I think "/proc/swaps" is not normal file and can not get EOF in some situation, so i use fgets a line and caculate swap dev number.

I guess the problem has been fixed by another patch [1], thus closing it in patchwork. Please let me know if not.

Kind regards,
Petr

[1] https://github.com/linux-test-project/ltp/commit/0f5d8c520d4e5b22325526eab41ed6bcad1847fc

Signature · 1 line
Mailing list info: https://lists.linux.it/listinfo/ltp
#3 lore raw
Quoted 7 lines · Petr Vorel

Hi Wei, Xu,

Quoted 4 lines

Hi Wei

Quoted 2 lines

I encounter a dead loop on following code in our test on ppc64 machine:
while ((c = fgetc(fp)) != EOF)

Quoted 7 lines
Quoted 1 line
Quoted 1 line

I think "/proc/swaps" is not normal file and can not get EOF in some

situation,

Quoted 1 line
Quoted 1 line

so i use fgets a line and caculate swap dev number.

I guess the problem has been fixed by another patch [1], thus closing it in patchwork. Please let me know if not.

Seems not, the patch [1] below is mainly to count the free disk size.

But Wei's work here is to calculate swap-dev numbers correctly (especially get rid of EOF affection to some degree).

Xu Yang, what do you think? or did I miss anything here?

Signature · 5 lines
Regards,
Li Wang

-- 
Mailing list info: https://lists.linux.it/listinfo/ltp
#4 lore raw
Quoted 13 lines · Petr Vorel
Quoted 11 lines

Hi Wei, Xu,

Quoted 6 lines

Hi Wei

Quoted 4 lines

I encounter a dead loop on following code in our test on ppc64 machine:
while ((c = fgetc(fp)) != EOF)

I think "/proc/swaps" is not normal file and can not get EOF in some

situation,

Quoted 1 line
Quoted 1 line

so i use fgets a line and caculate swap dev number.

Quoted 8 lines
Quoted 2 lines

I guess the problem has been fixed by another patch [1], thus closing it in patchwork. Please let me know if not.

Seems not, the patch [1] below is mainly to count the free disk size.

But Wei's work here is to calculate swap-dev numbers correctly (especially get rid of EOF affection to some degree).

OK, setting it as "New" again.

Kind regards,
Petr

Quoted 1 line

Xu Yang, what do you think? or did I miss anything here?

Signature · 1 line
Mailing list info: https://lists.linux.it/listinfo/ltp
#5 lore raw
In reply to Li Wang

Hi Li,

Quoted 22 lines

On Sun, Mar 3, 2024 at 9:14 PM Petr Vorel <[email protected] <mailto:[email protected]>> wrote:

Hi Wei, Xu,

> Hi  Wei

> > I encounter a dead loop on following code in our test on ppc64
machine:
> > while ((c = fgetc(fp)) != EOF)

> > I think "/proc/swaps" is not normal file and can not get EOF in
some situation,
> > so i use fgets a line and caculate swap dev number.

I guess the problem has been fixed by another patch [1], thus closing it in
patchwork. Please let me know if not.

Quoted 8 lines

Seems not, the patch [1] below is mainly to count the free disk size.

But Wei's work here is to calculate swap-dev numbers correctly (especially get rid of EOF affection to some degree).

Xu Yang, what do you think? or did I miss anything here?

I still think we can use the same way in ipc libs to get rid of the EOF problem instead of
searching "dev" keyword. We just don't need to calcualte "/proc/swaps" header.

int get_used_sysvipc(const char *file, const int lineno, const char *sysvipc_file)

{
	FILE *fp;
	int used = -1;
	char buf[BUFSIZE];

	fp = safe_fopen(file, lineno, NULL, sysvipc_file, "r");

	while (fgets(buf, BUFSIZE, fp) != NULL)
		used++;

	fclose(fp);

	if (used < 0) {
		tst_brk(TBROK, "can't read %s to get used sysvipc resource total at "
			"%s:%d", sysvipc_file, file, lineno);
	}

	return used;
}

But we don't get the actual reason, I still wonder why this deadloop exists o ppcc64 instead of
other architecture(ie x86_64).

Best Regards
Yang Xu

Quoted 4 lines
Signature · 3 lines
-- 
Regards,
Li Wang
Signature · 1 line
Mailing list info: https://lists.linux.it/listinfo/ltp
#6 lore raw
Quoted 29 lines · Yang Xu (Fujitsu)

Hi Li,

Quoted 27 lines

On Sun, Mar 3, 2024 at 9:14 PM Petr Vorel <[email protected] <mailto:[email protected]>> wrote:

Hi Wei, Xu,

> Hi Wei

> > I encounter a dead loop on following code in our test on ppc64
machine:
> > while ((c = fgetc(fp)) != EOF)

> > I think "/proc/swaps" is not normal file and can not get EOF in
some situation,
> > so i use fgets a line and caculate swap dev number.

I guess the problem has been fixed by another patch [1], thus closing it in
patchwork. Please let me know if not.

Seems not, the patch [1] below is mainly to count the free disk size.

But Wei's work here is to calculate swap-dev numbers correctly (especially get rid of EOF affection to some degree).

Quoted 9 lines
Quoted 3 lines

Xu Yang, what do you think? or did I miss anything here?

I still think we can use the same way in ipc libs to get rid of the EOF problem instead of
searching "dev" keyword. We just don't need to calcualte "/proc/swaps" header.

Sounds good to me. At least it counts right lines.

It'd be great to have a patch by that way.

Quoted 24 lines

int get_used_sysvipc(const char *file, const int lineno, const char *sysvipc_file)

{
        FILE *fp;
        int used = -1;
        char buf[BUFSIZE];

        fp = safe_fopen(file, lineno, NULL, sysvipc_file, "r");

        while (fgets(buf, BUFSIZE, fp) != NULL)
                used++;

        fclose(fp);

        if (used < 0) {
                tst_brk(TBROK, "can't read %s to get used sysvipc resource
total at "
                        "%s:%d", sysvipc_file, file, lineno);
        }

        return used;
}
Quoted 4 lines

But we don't get the actual reason, I still wonder why this deadloop exists o ppcc64 instead of
other architecture(ie x86_64).

Ok, I think we can just apply your suggested method to see if that deadloop disappears.

Signature · 5 lines
Regards,
Li Wang

-- 
Mailing list info: https://lists.linux.it/listinfo/ltp
#7 lore raw
Quoted 72 lines · Li Wang
Quoted 29 lines · Yang Xu (Fujitsu)

Hi Li,

Quoted 27 lines

On Sun, Mar 3, 2024 at 9:14 PM Petr Vorel <[email protected] <mailto:[email protected]>> wrote:

Hi Wei, Xu,

> Hi Wei

> > I encounter a dead loop on following code in our test on ppc64
machine:
> > while ((c = fgetc(fp)) != EOF)

> > I think "/proc/swaps" is not normal file and can not get EOF in
some situation,
> > so i use fgets a line and caculate swap dev number.

I guess the problem has been fixed by another patch [1], thus closing it in
patchwork. Please let me know if not.

Seems not, the patch [1] below is mainly to count the free disk size.

But Wei's work here is to calculate swap-dev numbers correctly (especially get rid of EOF affection to some degree).

Quoted 9 lines
Quoted 3 lines

Xu Yang, what do you think? or did I miss anything here?

I still think we can use the same way in ipc libs to get rid of the EOF problem instead of
searching "dev" keyword. We just don't need to calcualte "/proc/swaps" header.

Sounds good to me. At least it counts right lines.

It'd be great to have a patch by that way.

Quoted 24 lines

int get_used_sysvipc(const char *file, const int lineno, const char *sysvipc_file)

{
        FILE *fp;
        int used = -1;
        char buf[BUFSIZE];

        fp = safe_fopen(file, lineno, NULL, sysvipc_file, "r");

        while (fgets(buf, BUFSIZE, fp) != NULL)
                used++;

        fclose(fp);

        if (used < 0) {
                tst_brk(TBROK, "can't read %s to get used sysvipc resource
total at "
                        "%s:%d", sysvipc_file, file, lineno);
        }

        return used;
}
Quoted 8 lines
Quoted 4 lines

But we don't get the actual reason, I still wonder why this deadloop exists o ppcc64 instead of
other architecture(ie x86_64).

Ok, I think we can just apply your suggested method to see if that deadloop disappears.

Hi Petr, Xu, Li
Deadloop will disappear if you use fgets, fgetc can not get EOF on ppc64.(Suspect an bug) So either use my patch or Xu's suggestion both can work.

Quoted 8 lines
Signature · 6 lines
-- 
Regards,
Li Wang

-- 
Mailing list info: https://lists.linux.it/listinfo/ltp
Signature · 1 line
Mailing list info: https://lists.linux.it/listinfo/ltp
#8 lore raw
Quoted 70 lines · Li Wang
Quoted 68 lines
Quoted 29 lines · Yang Xu (Fujitsu)

Hi Li,

Quoted 24 lines

On Sun, Mar 3, 2024 at 9:14 PM Petr Vorel <[email protected] <mailto:[email protected]>> wrote:

Hi Wei, Xu,

> Hi Wei

> > I encounter a dead loop on following code in our test on ppc64
machine:
> > while ((c = fgetc(fp)) != EOF)

> > I think "/proc/swaps" is not normal file and can not get EOF in
some situation,
> > so i use fgets a line and caculate swap dev number.

I guess the problem has been fixed by another patch [1], thus closing it in
patchwork. Please let me know if not.

Seems not, the patch [1] below is mainly to count the free disk size.

But Wei's work here is to calculate swap-dev numbers correctly (especially get rid of EOF affection to some degree).

Quoted 8 lines
Quoted 1 line

Xu Yang, what do you think? or did I miss anything here?

I still think we can use the same way in ipc libs to get rid of the EOF problem instead of
searching "dev" keyword. We just don't need to calcualte "/proc/swaps" header.

Sounds good to me. At least it counts right lines.

It'd be great to have a patch by that way.

Quoted 19 lines

int get_used_sysvipc(const char *file, const int lineno, const char *sysvipc_file)

{
        FILE *fp;
        int used = -1;
        char buf[BUFSIZE];

        fp = safe_fopen(file, lineno, NULL, sysvipc_file, "r");

        while (fgets(buf, BUFSIZE, fp) != NULL)
                used++;

        fclose(fp);

        if (used < 0) {
                tst_brk(TBROK, "can't read %s to get used sysvipc resource
total at "
                        "%s:%d", sysvipc_file, file, lineno);
        }
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
        return used;
}

But we don't get the actual reason, I still wonder why this deadloop exists o ppcc64 instead of
other architecture(ie x86_64).

FYI also other architectures fails (eg.g least aarch64, s390x and ppc64le which was in the original report; it looks to me it works on x86_64).

Problem is somewhere in c1b8c011e447088b08787462e0c2ed50cd8c43f3..8fd941649afeecc5f87508c9f94e9a840a84e44d which contains 319693d0b ("libltpswap: alter tst_count_swaps API"), which is Wei trying to fix.

Quoted 6 lines
Quoted 2 lines

Ok, I think we can just apply your suggested method to see if that deadloop disappears.

Hi Petr, Xu, Li
Deadloop will disappear if you use fgets, fgetc can not get EOF on ppc64.(Suspect an bug) So either use my patch or Xu's suggestion both can work.

I would vote for Xu's suggestion (any line except the first header is valid), actually this implementation is not correct when it counts only /dev:

# dd if=/dev/zero of=/root/swap bs=100M count=1
# mkswap -f /root/swap
# swapon /root/swap
# cat /proc/swaps
Filename				Type		Size		Used		Priority
/dev/dm-0                               partition	192508		4048		-2
/root/swap                              file		102396		0		-3

(second line without /dev)

Also typo in the subject: s/caculate/calculate/

Wei, please send new version and Cc us.

Kind regards,
Petr

Signature · 1 line
Mailing list info: https://lists.linux.it/listinfo/ltp
#9 lore raw

PATCH v2 1/1 libswap.c: Improve calculate swap dev number

In reply to Wei Gao via ltp

I encounter a dead loop on following code in our test on ppc64 machine:
while ((c = fgetc(fp)) != EOF)

Use fgets instead of fgetc can avoid above issue.

---
 libs/libltpswap/libswap.c | 9 ++++-----
 1 file changed, 4 insertions(+), 5 deletions(-)
libs/libltpswap/libswap.c +4 -5
index c10db91c0..a26ea25e4 100644
@@ -13,6 +13,7 @@
 
 #define TST_NO_DEFAULT_MAIN
 #define DEFAULT_MAX_SWAPFILE 32
+#define BUFSIZE 200
 
 #include "tst_test.h"
 #include "libswap.h"
@@ -279,16 +280,14 @@ int tst_count_swaps(void)
 {
 	FILE *fp;
 	int used = -1;
-	char c;
+	char buf[BUFSIZE];
 
 	fp = SAFE_FOPEN("/proc/swaps", "r");
 	if (fp == NULL)
 		return -1;
 
-	while ((c = fgetc(fp)) != EOF) {
-		if (c == '\n')
-			used++;
-	}
+	while (fgets(buf, BUFSIZE, fp) != NULL)
+		used++;
 
 	SAFE_FCLOSE(fp);
 	if (used < 0)
Signature · 5 lines
2.35.3


-- 
Mailing list info: https://lists.linux.it/listinfo/ltp
Reviewed-by on v2 1/1
#10 lore raw

Re: [LTP] [PATCH v2] libswap.c: Improve calculate swap dev number

Quoted 2 lines

I encounter a dead loop on following code in our test on ppc64 machine:
while ((c = fgetc(fp)) != EOF)

FYI it fails also on aarch64 and s390x, thus I'll add it to the commit message before merge. I also try to have look why it's broken.

Quoted 1 line

Use fgets instead of fgetc can avoid above issue.

LGTM. I probably merge it tonight.

Kind regards,
Petr

Quoted 37 lines
---
 libs/libltpswap/libswap.c | 9 ++++-----
 1 file changed, 4 insertions(+), 5 deletions(-)
libs/libltpswap/libswap.c +4 -5
index c10db91c0..a26ea25e4 100644
@@ -13,6 +13,7 @@
 
 #define TST_NO_DEFAULT_MAIN
 #define DEFAULT_MAX_SWAPFILE 32
+#define BUFSIZE 200
 
 #include "tst_test.h"
 #include "libswap.h"
@@ -279,16 +280,14 @@ int tst_count_swaps(void)
 {
 	FILE *fp;
 	int used = -1;
-	char c;
+	char buf[BUFSIZE];
 
 	fp = SAFE_FOPEN("/proc/swaps", "r");
 	if (fp == NULL)
 		return -1;
 
-	while ((c = fgetc(fp)) != EOF) {
-		if (c == '\n')
-			used++;
-	}
+	while (fgets(buf, BUFSIZE, fp) != NULL)
+		used++;
 
 	SAFE_FCLOSE(fp);
 	if (used < 0)
Signature · 1 line
Mailing list info: https://lists.linux.it/listinfo/ltp
#11 lore raw

Re: [LTP] [PATCH v2] libswap.c: Improve calculate swap dev number

In reply to Wei Gao via ltp

Hi Wei, all,

thanks, merged!

FYI swapon03 with this patch still fails on SLES 12-SP5 (4.12 based kernel):

tst_kconfig.c:87: TINFO: Parsing kernel config '/proc/config.gz' tst_kconfig.c:87: TINFO: Parsing kernel config '/proc/config.gz' swapon03.c:55: TFAIL: swapon(filename, 0) failed: EPERM (1) swapon03.c:55: TFAIL: swapon(filename, 0) failed: EPERM (1)

But that's likely fails even without the patch (going to verify older SLES releases).

Kind regards,
Petr

Signature · 1 line
Mailing list info: https://lists.linux.it/listinfo/ltp
#12 lore raw

Re: [LTP] [PATCH v2] libswap.c: Improve calculate swap dev number

Hi Petr, all,

Quoted 11 lines

Hi Wei, all,

thanks, merged!

FYI swapon03 with this patch still fails on SLES 12-SP5 (4.12 based kernel):

tst_kconfig.c:87: TINFO: Parsing kernel config '/proc/config.gz' tst_kconfig.c:87: TINFO: Parsing kernel config '/proc/config.gz' swapon03.c:55: TFAIL: swapon(filename, 0) failed: EPERM (1) swapon03.c:55: TFAIL: swapon(filename, 0) failed: EPERM (1)

Quoted 5 lines

But that's likely fails even without the patch (going to verify older SLES releases).

Kind regards,
Petr

I guess SLES 12-SP5 has backported some patch that modified kernel constant MAX_SWAPFILES value[1]. But ltp doesn't add fallback for SLES like I did for RHEL9.

I think you can paste your kernel code(include/linux/swap.h), then you or we can figure out what cause this failure.

[1]https://github.com/linux-test-project/ltp/commit/c1b8c011e447088b08787462e0c2ed50cd8c43f3

Best Regards
Yang Xu

Signature · 1 line
Mailing list info: https://lists.linux.it/listinfo/ltp
#13 lore raw

Re: [LTP] [PATCH v2] libswap.c: Improve calculate swap dev number

Hi Xu,

Quoted 17 lines

Hi Petr, all,

Quoted 13 lines

Hi Wei, all,

thanks, merged!

FYI swapon03 with this patch still fails on SLES 12-SP5 (4.12 based kernel):

tst_kconfig.c:87: TINFO: Parsing kernel config '/proc/config.gz' tst_kconfig.c:87: TINFO: Parsing kernel config '/proc/config.gz' swapon03.c:55: TFAIL: swapon(filename, 0) failed: EPERM (1) swapon03.c:55: TFAIL: swapon(filename, 0) failed: EPERM (1)

But that's likely fails even without the patch (going to verify older SLES releases).

Quoted 6 lines
Quoted 2 lines

Kind regards,
Petr

I guess SLES 12-SP5 has backported some patch that modified kernel constant MAX_SWAPFILES value[1]. But ltp doesn't add fallback for SLES like I did for RHEL9.

Very good point, there are some patches.

Quoted 2 lines

I think you can paste your kernel code(include/linux/swap.h), then you or we can figure out what cause this failure.

I'll try to figure out what is missing and post the config (either with patches or just config). Thanks!

Quoted 4 lines

[1]https://github.com/linux-test-project/ltp/commit/c1b8c011e447088b08787462e0c2ed50cd8c43f3

Best Regards
Yang Xu

Signature · 1 line
Mailing list info: https://lists.linux.it/listinfo/ltp