linux-kernel - Re: next-20080808: bash: fork: Resource temporarily unavailable

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Sun, 10 Aug 2008 10:45:39 +0100
From:	David Howells <dhowells@...hat.com>
To:	Jasper Bryant-Greene <jasper@...ton.co.nz>
Cc:	dhowells@...hat.com, Thomas Meyer <thomas@...3r.de>,
	James Morris <jmorris@...ei.org>,
	Linux-Kernel <linux-kernel@...r.kernel.org>,
	Linux-Next <linux-next@...r.kernel.org>
Subject: Re: next-20080808: bash: fork: Resource temporarily unavailable


Jasper Bryant-Greene <jasper@...ton.co.nz> wrote:

> This occurs on my x86_64 system, config was posted to this thread
> earlier.

Does it happen every time?  I took your config modified it very slightly so
that it'd work on my system:

	-CONFIG_INITRAMFS_SOURCE="/usr/share/v86d/initramfs"
	-CONFIG_INITRAMFS_ROOT_UID=0
	-CONFIG_INITRAMFS_ROOT_GID=0
	+CONFIG_INITRAMFS_SOURCE=""

	-CONFIG_PACKET=m
	+CONFIG_PACKET=y

	-CONFIG_UNIX=m
	+CONFIG_UNIX=y

	-CONFIG_NET_KEY=m
	+CONFIG_NET_KEY=y

	-CONFIG_E1000E=m
	+CONFIG_E1000E=y

	-# CONFIG_SERIAL_8250 is not set
	+CONFIG_SERIAL_8250=y
	+CONFIG_SERIAL_8250_CONSOLE=y
	+CONFIG_SERIAL_8250_PCI=y
	+CONFIG_SERIAL_8250_PNP=y
	+CONFIG_SERIAL_8250_NR_UARTS=4
	+CONFIG_SERIAL_8250_RUNTIME_UARTS=4
	+# CONFIG_SERIAL_8250_EXTENDED is not set

	+CONFIG_SERIAL_CORE=y
	+CONFIG_SERIAL_CORE_CONSOLE=y

	-# CONFIG_EXT3_FS is not set
	+CONFIG_EXT3_FS=y
	+CONFIG_EXT3_FS_XATTR=y
	+CONFIG_EXT3_FS_POSIX_ACL=y
	+CONFIG_EXT3_FS_SECURITY=y
	+CONFIG_JBD=y
	+# CONFIG_JBD_DEBUG is not set

None of which should make a difference, but the kernel works fine and survives
the LTP syscalls testscript:-/

Hmmm... There is a possibility, I suppose.  The only place in the fork path
that I know of that returns EAGAIN is this bit:

	retval = -EAGAIN;
	if (atomic_read(&p->real_cred->user->processes) >=
			p->signal->rlim[RLIMIT_NPROC].rlim_cur) {
		if (!capable(CAP_SYS_ADMIN) && !capable(CAP_SYS_RESOURCE) &&
		    p->real_cred->user != current->nsproxy->user_ns->root_user)
			goto bad_fork_free;
	}

Are you running as an unprivileged user?  If so, this may be the point that's
biting you.  Can you try applying the attached patch to find more information?

David
---
Add printks to find problem

From: David Howells <dhowells@...hat.com>


---

 kernel/fork.c |   16 +++++++++++++---
 1 files changed, 13 insertions(+), 3 deletions(-)


diff --git a/kernel/fork.c b/kernel/fork.c
index eb7838d..4d52954 100644
--- a/kernel/fork.c
+++ b/kernel/fork.c
@@ -920,8 +920,10 @@ static struct task_struct *copy_process(unsigned long clone_flags,
 		return ERR_PTR(-EINVAL);
 
 	retval = security_task_create(clone_flags);
-	if (retval)
+	if (retval) {
+		printk(KERN_ERR "security_task_create() returned %d\n", retval);
 		goto fork_out;
+	}
 
 	retval = -ENOMEM;
 	p = dup_task_struct(current);
@@ -938,13 +940,20 @@ static struct task_struct *copy_process(unsigned long clone_flags,
 	if (atomic_read(&p->real_cred->user->processes) >=
 			p->signal->rlim[RLIMIT_NPROC].rlim_cur) {
 		if (!capable(CAP_SYS_ADMIN) && !capable(CAP_SYS_RESOURCE) &&
-		    p->real_cred->user != current->nsproxy->user_ns->root_user)
+		    p->real_cred->user != current->nsproxy->user_ns->root_user) {
+			printk(KERN_ERR "Rlimit EAGAIN (%d >= %lu, uid %d)\n",
+			       atomic_read(&p->real_cred->user->processes),
+			       p->signal->rlim[RLIMIT_NPROC].rlim_cur,
+			       p->real_cred->user->uid);
 			goto bad_fork_free;
+		}
 	}
 
 	retval = copy_creds(p, clone_flags);
-	if (retval < 0)
+	if (retval < 0) {
+		printk(KERN_ERR "copy_creds() returned %d\n", retval);
 		goto bad_fork_free;
+	}
 
 	/*
 	 * If multiple threads are within copy_process(), then this check
@@ -1274,6 +1283,7 @@ bad_fork_cleanup_count:
 bad_fork_free:
 	free_task(p);
 fork_out:
+	printk(KERN_ERR "copy_process() = %d\n", retval);
 	return ERR_PTR(retval);
 }
 
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/