Re: Fwd: [musl] Re: musl getaddr info breakage on older kernels

[Satadru Pramanik <satadru@gmail.com>]

[-- Attachment #1.1: Type: text/plain, Size: 3157 bytes --]

I have attached the working patch I have ended up using, which has added
only a definition for ENOSYS.

Thanks again. It's nice to be offering software for these machines with
working, current libc. :)

Satadru

On Thu, Feb 17, 2022 at 5:48 PM Rich Felker <dalias@libc.org> wrote:

> On Thu, Feb 17, 2022 at 04:43:51PM -0500, Satadru Pramanik wrote:
> > > issues in userspace to maintain some functionality for any users who
> may
> > > > still be using the device.
> > > >
> > > > The simplest workaround possible would be ideal.
> > >
> > > If you're shipping binaries specifically for these devices, the
> > > simplest fix is just to emulate the failure that should happen in the
> > > kernel in userspace, using the attached patch. DO NOT deploy this
> > > patch in binaries meant to be used on modern systems, since they will
> > > break when Y2038 rolls around. (Your old Chromebooks will break then
> > > too.)
> > >
> > I'll let the next generation of preservationists worry about the Y2038
> > issues. Thanks for the patch. I'll build that now with the
> >
> https://git.musl-libc.org/cgit/musl/patch/?id=c2feda4e2ea61f4da73f2f38b2be5e327a7d1a91
> > reversal patch for the i686 machines, and see if that fixes the issue. If
>
> That reversal should no longer be needed with the other patch, which
> is a much bigger hammer. Reversing that just got rid of using the new
> socket-related syscalls; the new hack patch gets rid of using all new
> syscalls.
>
> > so, we can consider the matter closed, unless you want to suggest a
> > modification of the syscall patch to handle older working kernels which
> > might avoid the need for the patch when used with older systems.
>
> The patch is fine for older broken or older working systems. It just
> suppresses the ability to use any syscall added after Linux 3.8, so
> it's a bad idea to use on *newer* systems.
>
> > > It is interesting though
> > > > that the sample program works fine when built against near-stock
> glibc
> > > > 2.23, no?
> > >
> > > No. If your kernel has a bug that makes something behave wildly wrong,
> > > whether you do or don't see that as visible breakage with a particular
> > > piece of software is not particularly interesting.
> > >
> > > I'm pretty sure, however, that you just haven't tested enough to see
> > > any failures. glibc 2.23 is from 2016, so any functionality in it
> > > using syscalls added after 2011 (3.8 kernel) is going to blow up
> > > badly, thinking the syscall succeeded and returned some positive value
> > > when actually the kernel lacks it.
> > >
> > > In the particular case of clock_gettime, it's just that your glibc
> > > 2.23 has a hard Y2038 EOL and does not use/support the missing time64
> > > syscalls.
> > >
> > >
> > Duly noted.
> >
> > Thanks so much for being patient while I got you enough information to
> fix
> > this!
> >
> > On behalf of the entire Chromebrew community, thank you!
>
> And thanks for reporting, running tests, and following through on
> this. It will be useful to know this is an issue others might hit, and
> to be able to check other old systems that might have backported the
> patch with the bug.
>

[-- Attachment #1.2: Type: text/html, Size: 4070 bytes --]

[-- Attachment #2: broken_chromeos_kernel_hack_2.diff --]
[-- Type: application/octet-stream, Size: 2653 bytes --]

--- a/arch/i386/syscall_arch.h	2022-02-17 16:45:37.398583011 -0500
+++ b/arch/i386/syscall_arch.h	2022-02-17 16:50:02.311265598 -0500
@@ -9,11 +9,21 @@
 #define SYSCALL_INSNS "call *%%gs:16"
 #endif
 
+/*
+ *  * This error code is special: arch syscall entry code will return
+ *   * -ENOSYS if users try to call a syscall that doesn't exist.  To keep
+ *    * failures of syscalls that really do exist distinguishable from
+ *     * failures due to attempts to use a nonexistent syscall, syscall
+ *      * implementations should refrain from returning -ENOSYS.
+ *       */
+#define ENOSYS          38      /* Invalid system call number */
+
 #define SYSCALL_INSNS_12 "xchg %%ebx,%%edx ; " SYSCALL_INSNS " ; xchg %%ebx,%%edx"
 #define SYSCALL_INSNS_34 "xchg %%ebx,%%edi ; " SYSCALL_INSNS " ; xchg %%ebx,%%edi"
 
 static inline long __syscall0(long n)
 {
+	if (n>350) return -ENOSYS;
 	unsigned long __ret;
 	__asm__ __volatile__ (SYSCALL_INSNS : "=a"(__ret) : "a"(n) : "memory");
 	return __ret;
@@ -21,6 +31,7 @@
 
 static inline long __syscall1(long n, long a1)
 {
+	if (n>350) return -ENOSYS;
 	unsigned long __ret;
 	__asm__ __volatile__ (SYSCALL_INSNS_12 : "=a"(__ret) : "a"(n), "d"(a1) : "memory");
 	return __ret;
@@ -28,6 +39,7 @@
 
 static inline long __syscall2(long n, long a1, long a2)
 {
+	if (n>350) return -ENOSYS;
 	unsigned long __ret;
 	__asm__ __volatile__ (SYSCALL_INSNS_12 : "=a"(__ret) : "a"(n), "d"(a1), "c"(a2) : "memory");
 	return __ret;
@@ -35,6 +47,7 @@
 
 static inline long __syscall3(long n, long a1, long a2, long a3)
 {
+	if (n>350) return -ENOSYS;
 	unsigned long __ret;
 #if !defined(__PIC__) || !defined(BROKEN_EBX_ASM)
 	__asm__ __volatile__ (SYSCALL_INSNS : "=a"(__ret) : "a"(n), "b"(a1), "c"(a2), "d"(a3) : "memory");
@@ -46,6 +59,7 @@
 
 static inline long __syscall4(long n, long a1, long a2, long a3, long a4)
 {
+	if (n>350) return -ENOSYS;
 	unsigned long __ret;
 #if !defined(__PIC__) || !defined(BROKEN_EBX_ASM)
 	__asm__ __volatile__ (SYSCALL_INSNS : "=a"(__ret) : "a"(n), "b"(a1), "c"(a2), "d"(a3), "S"(a4) : "memory");
@@ -57,6 +71,7 @@
 
 static inline long __syscall5(long n, long a1, long a2, long a3, long a4, long a5)
 {
+	if (n>350) return -ENOSYS;
 	unsigned long __ret;
 #if !defined(__PIC__) || !defined(BROKEN_EBX_ASM)
 	__asm__ __volatile__ (SYSCALL_INSNS
@@ -70,6 +85,7 @@
 
 static inline long __syscall6(long n, long a1, long a2, long a3, long a4, long a5, long a6)
 {
+	if (n>350) return -ENOSYS;
 	unsigned long __ret;
 #if !defined(__PIC__) || !defined(BROKEN_EBX_ASM)
 	__asm__ __volatile__ ("pushl %7 ; push %%ebp ; mov 4(%%esp),%%ebp ; " SYSCALL_INSNS " ; pop %%ebp ; add $4,%%esp"