um: Remove dead usage of TIF_IA32

This seems like a dead artifact since TIF_IA32 is not even defined as a TI flag for UM. Looking back in git history, it made sense in the old days, but it is apparently not used since UM was split out of the x86 arch/. It is also going away from the x86 tree soon.

Also, I think the variable clean up it performs is not needed as 64-bit UML doesn't run 32-bit binaries as far as I can tell, and 32-bit UML has 32-bit ulong.

---
 arch/x86/um/ptrace_64.c | 13 +------------
 1 file changed, 1 insertion(+), 12 deletions(-)

index 09a085bde0d4..1401899dee9b 100644
@@ -52,14 +52,6 @@ static const int reg_offsets[] =
 
 int putreg(struct task_struct *child, int regno, unsigned long value)
 {
-#ifdef TIF_IA32
-	/*
-	 * Some code in the 64bit emulation may not be 64bit clean.
-	 * Don't take any chances.
-	 */
-	if (test_tsk_thread_flag(child, TIF_IA32))
-		value &= 0xffffffff;
-#endif
 	switch (regno) {
 	case R8:
 	case R9:

@@ -137,10 +129,7 @@ int poke_user(struct task_struct *child, long addr, long data)
 unsigned long getreg(struct task_struct *child, int regno)
 {
 	unsigned long mask = ~0UL;
-#ifdef TIF_IA32
-	if (test_tsk_thread_flag(child, TIF_IA32))
-		mask = 0xffffffff;
-#endif
+
 	switch (regno) {
 	case R8:
 	case R9:

2.28.0


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

Yep, that's true. I was looking into it once _very_ briefly, but no :)

johannes

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

----- Ursprüngliche Mail -----

Yep. UML has no IA32 emulation and never had.
Looks like a copy&paste leftover.

Thanks,
//richard

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