New kernels

[musa]

Hi
vmlinux-6.10-12 got a kernel error

[xeno74]

Hi
vmlinux-6.10-12 got a kernel error

Thank you for testing. I have created a patch for reverting the first bad commit.

For the RC3:

unix_diag.patch:

Code: Select all

--- b/net/unix/diag.c	2024-06-24 21:28:33.951978761 +0200
+++ a/net/unix/diag.c	2024-06-24 21:30:36.767978293 +0200
@@ -65,7 +65,7 @@ static int sk_diag_dump_icons(struct soc
 	u32 *buf;
 	int i;
 
-	if (READ_ONCE(sk->sk_state) == TCP_LISTEN) {
+	if (sk->sk_state == TCP_LISTEN) {
 		spin_lock(&sk->sk_receive_queue.lock);
 
 		attr = nla_reserve(nlskb, UNIX_DIAG_ICONS,
@@ -103,7 +103,7 @@ static int sk_diag_show_rqlen(struct soc
 {
 	struct unix_diag_rqlen rql;
 
-	if (READ_ONCE(sk->sk_state) == TCP_LISTEN) {
+	if (sk->sk_state == TCP_LISTEN) {
 		rql.udiag_rqueue = sk->sk_receive_queue.qlen;
 		rql.udiag_wqueue = sk->sk_max_ack_backlog;
 	} else {
@@ -136,7 +136,7 @@ static int sk_diag_fill(struct sock *sk,
 	rep = nlmsg_data(nlh);
 	rep->udiag_family = AF_UNIX;
 	rep->udiag_type = sk->sk_type;
-	rep->udiag_state = READ_ONCE(sk->sk_state);
+	rep->udiag_state = sk->sk_state;
 	rep->pad = 0;
 	rep->udiag_ino = sk_ino;
 	sock_diag_save_cookie(sk, rep->udiag_cookie);
@@ -215,7 +215,7 @@ static int unix_diag_dump(struct sk_buff
 		sk_for_each(sk, &net->unx.table.buckets[slot]) {
 			if (num < s_num)
 				goto next;
-			if (!(req->udiag_states & (1 << READ_ONCE(sk->sk_state))))
+			if (!(req->udiag_states & (1 << sk->sk_state)))
 				goto next;
 			if (sk_diag_dump(sk, skb, req, sk_user_ns(skb->sk),
 					 NETLINK_CB(cb->skb).portid,

For the RC5:

unix_diag_v2.patch:

Code: Select all

--- b/net/unix/diag.c	2024-06-23 23:08:54.000000000 +0200
+++ a/net/unix/diag.c	2024-06-24 22:22:57.959966310 +0200
@@ -65,7 +65,7 @@ static int sk_diag_dump_icons(struct soc
 	u32 *buf;
 	int i;
 
-	if (READ_ONCE(sk->sk_state) == TCP_LISTEN) {
+	if (sk->sk_state == TCP_LISTEN) {	
 		spin_lock(&sk->sk_receive_queue.lock);
 
 		attr = nla_reserve(nlskb, UNIX_DIAG_ICONS,
@@ -103,7 +103,7 @@ static int sk_diag_show_rqlen(struct soc
 {
 	struct unix_diag_rqlen rql;
 
-	if (READ_ONCE(sk->sk_state) == TCP_LISTEN) {
+	if (sk->sk_state == TCP_LISTEN) {	
 		rql.udiag_rqueue = skb_queue_len_lockless(&sk->sk_receive_queue);
 		rql.udiag_wqueue = sk->sk_max_ack_backlog;
 	} else {
@@ -136,7 +136,7 @@ static int sk_diag_fill(struct sock *sk,
 	rep = nlmsg_data(nlh);
 	rep->udiag_family = AF_UNIX;
 	rep->udiag_type = sk->sk_type;
-	rep->udiag_state = READ_ONCE(sk->sk_state);
+	rep->udiag_state = sk->sk_state;	
 	rep->pad = 0;
 	rep->udiag_ino = sk_ino;
 	sock_diag_save_cookie(sk, rep->udiag_cookie);
@@ -215,7 +215,7 @@ static int unix_diag_dump(struct sk_buff
 		sk_for_each(sk, &net->unx.table.buckets[slot]) {
 			if (num < s_num)
 				goto next;
-			if (!(req->udiag_states & (1 << READ_ONCE(sk->sk_state))))
+			if (!(req->udiag_states & (1 << sk->sk_state)))	
 				goto next;
 			if (sk_diag_dump(sk, skb, req, sk_user_ns(skb->sk),
 					 NETLINK_CB(cb->skb).portid,

I compiled a new test kernel with this patch.

Dowload: vmlinux-6.10-13

Please test it.

Thanks,
Christian

[Hypex]

Could you also please test the kernel vmlinux-6.10-10?

Thanks for info. Okay so tested this after posting and same lightdm error. Will keep testing following kernels.

[xeno74]

Thanks for info. Okay so tested this after posting and same lightdm error. Will keep testing following kernels.

Thanks a lot for testing!

I think you don’t need to test vmlinux-6.10-11 and vmlinux-6.10-12 because Musa has already tested them.

Could you please test the kernel vmlinux-6.10-13?

This kernel is already patched.

[musa]

hi
vmlinux-6.10-13 got a kernel panic

[xeno74]

hi
vmlinux-6.10-13 got a kernel panic

OK, please test vmlinux-6.10-1 again.

[xeno74]

Musa,

You had the same issue in 2021 and solved it.

Hi Christian
I have tried to boot kernel 5.12 rc7 but the kernel can not identify sata 4 (my linux hard drive).
Have new formatted my usbdisk with Fat 16. no luck. When I afterwards start the same partition with kernel 5.11 all errors are gone. Therefore, I believe that it is the kernel that is the problem.
Have a nice day
Peter

Link: Kernel 5.12

How did you solve this issue?

Maybe an issue with the plugs from the cable into the hard drive?

[musa]

Hi
In the next 2 weeks I will not be at home much and therefore will not be able to test very much. I
have tested vmlinux-6.10-1 again and the same result: kernel panic.

I don't think there is a connection with something mechanical in my machine and the fact that some kernels run and others get kernel panic.
I think I sometimes have problems with corrupt fat file systems and or kernels that are damaged when copying or unpacking.

All my 9 linux distros boot on 5.11 perfectly stable.
There are only problems with 5.11 if I copy too many files over to CF0: Then I get boot problems and have to reformat cf0: and only have a few files copied over. After that there are no problems with 5.11.
Sometimes I get kernel panic on a new kernel and have been successful in downloading, extracting and copying to usb disk on another of my computers, after which the kernel boots. Other times it doesn't change anything, like now. And I don't now why.
ps

I have tested several of the kernels 5.10-1 > 5.10-13 with the same result.
The bad ones are still bad ones and the good ones are still good
Have a nice day.

[xeno74]

Hi Musa,

Many thanks for all the tests. You did your best. Well done!
I think the kernel support for the X1000 is broken.

I don’t have any ideas anymore currently.

Maybe we can release some stable kernels in the future.

Thanks,
Christian

[Hypex]

Thanks a lot for testing!

I think you don’t need to test vmlinux-6.10-11 and vmlinux-6.10-12 because Musa has already tested them.

Could you please test the kernel vmlinux-6.10-13?

This kernel is already patched.

I suppose I like to give them all a test, even though I have had missed a few.

• vmlinux-6.10-1: panic
• vmlinux-6.10-10: good
• vmlinux-6.10-11: good
• vmlinux-6.10-12: panic
• vmlinux-6.10-13: panic

The good ones can boot to a prompt with lightdm error as before. But I can log in on terminal. All give sata errors and have trouble setting up ports.

For some reason if I boot kernel 5.10.218 it also has sata issues. But it can boot. Typing this on Precise now. Might need to sort my names better so I know what's the newest.