um: Rework syscall handling

Rework syscall handling to be platform independent. Also create a clean
split between queueing of syscalls and flushing them out, removing the
need to keep state in the code that triggers the syscalls.

The code adds syscall_data_len to the global mm_id structure. This will
be used later to allow surrounding code to track whether syscalls still
need to run and if errors occurred.

Signed-off-by: Benjamin Berg <benjamin@sipsolutions.net>
Link: https://patch.msgid.link/20240703134536.1161108-5-benjamin@sipsolutions.net
Signed-off-by: Johannes Berg <johannes.berg@intel.com>

5 files changed, 16 insertions, 134 deletions

diff --git a/arch/x86/um/Makefile b/arch/x86/um/Makefile
index 8bc72a51b257..4559d2a693c4 100644
--- a/arch/x86/um/Makefile
+++ b/arch/x86/um/Makefile
@@ -11,7 +11,7 @@ endif
 
 obj-y = bugs_$(BITS).o delay.o fault.o ldt.o \
 	ptrace_$(BITS).o ptrace_user.o setjmp_$(BITS).o signal.o \
-	stub_$(BITS).o stub_segv.o \
+	stub_segv.o \
 	sys_call_table_$(BITS).o sysrq_$(BITS).o tls_$(BITS).o \
 	mem_$(BITS).o subarch.o os-Linux/
 
diff --git a/arch/x86/um/ldt.c b/arch/x86/um/ldt.c
index 255a44dd415a..c99fc23b1290 100644
--- a/arch/x86/um/ldt.c
+++ b/arch/x86/um/ldt.c
@@ -12,33 +12,22 @@
 #include <os.h>
 #include <skas.h>
 #include <sysdep/tls.h>
+#include <stub-data.h>
 
 static inline int modify_ldt (int func, void *ptr, unsigned long bytecount)
 {
 	return syscall(__NR_modify_ldt, func, ptr, bytecount);
 }
 
-static long write_ldt_entry(struct mm_id *mm_idp, int func,
-		     struct user_desc *desc, void **addr, int done)
+static void write_ldt_entry(struct mm_id *mm_idp, int func,
+		     struct user_desc *desc)
 {
-	long res;
-	void *stub_addr;
-
-	BUILD_BUG_ON(sizeof(*desc) % sizeof(long));
-
-	res = syscall_stub_data(mm_idp, (unsigned long *)desc,
-				sizeof(*desc) / sizeof(long),
-				addr, &stub_addr);
-	if (!res) {
-		unsigned long args[] = { func,
-					 (unsigned long)stub_addr,
-					 sizeof(*desc),
-					 0, 0, 0 };
-		res = run_syscall_stub(mm_idp, __NR_modify_ldt, args,
-				       0, addr, done);
-	}
+	struct stub_syscall *sc;
 
-	return res;
+	sc = syscall_stub_alloc(mm_idp);
+	sc->syscall = STUB_SYSCALL_LDT;
+	sc->ldt.func = func;
+	memcpy(&sc->ldt.desc, desc, sizeof(*desc));
 }
 
 /*
@@ -127,7 +116,6 @@ static int write_ldt(void __user * ptr, unsigned long bytecount, int func)
 	int i, err;
 	struct user_desc ldt_info;
 	struct ldt_entry entry0, *ldt_p;
-	void *addr = NULL;
 
 	err = -EINVAL;
 	if (bytecount != sizeof(ldt_info))
@@ -148,7 +136,8 @@ static int write_ldt(void __user * ptr, unsigned long bytecount, int func)
 
 	mutex_lock(&ldt->lock);
 
-	err = write_ldt_entry(mm_idp, func, &ldt_info, &addr, 1);
+	write_ldt_entry(mm_idp, func, &ldt_info);
+	err = syscall_stub_flush(mm_idp);
 	if (err)
 		goto out_unlock;
 
@@ -166,7 +155,8 @@ static int write_ldt(void __user * ptr, unsigned long bytecount, int func)
 				err = -ENOMEM;
 				/* Undo the change in host */
 				memset(&ldt_info, 0, sizeof(ldt_info));
-				write_ldt_entry(mm_idp, 1, &ldt_info, &addr, 1);
+				write_ldt_entry(mm_idp, 1, &ldt_info);
+				err = syscall_stub_flush(mm_idp);
 				goto out_unlock;
 			}
 			if (i == 0) {
@@ -303,7 +293,6 @@ long init_new_ldt(struct mm_context *new_mm, struct mm_context *from_mm)
 	short * num_p;
 	int i;
 	long page, err=0;
-	void *addr = NULL;
 
 
 	mutex_init(&new_mm->arch.ldt.lock);
@@ -318,11 +307,9 @@ long init_new_ldt(struct mm_context *new_mm, struct mm_context *from_mm)
 		ldt_get_host_info();
 		for (num_p=host_ldt_entries; *num_p != -1; num_p++) {
 			desc.entry_number = *num_p;
-			err = write_ldt_entry(&new_mm->id, 1, &desc,
-					      &addr, *(num_p + 1) == -1);
-			if (err)
-				break;
+			write_ldt_entry(&new_mm->id, 1, &desc);
 		}
+		err = syscall_stub_flush(&new_mm->id);
 		new_mm->arch.ldt.entry_count = 0;
 
 		goto out;
diff --git a/arch/x86/um/shared/sysdep/stub.h b/arch/x86/um/shared/sysdep/stub.h
index ce0ca46ad383..579681d12158 100644
--- a/arch/x86/um/shared/sysdep/stub.h
+++ b/arch/x86/um/shared/sysdep/stub.h
@@ -12,4 +12,5 @@
 #endif
 
 extern void stub_segv_handler(int, siginfo_t *, void *);
+extern void stub_syscall_handler(void);
 extern void stub_clone_handler(void);
diff --git a/arch/x86/um/stub_32.S b/arch/x86/um/stub_32.S
deleted file mode 100644
index 8291899e6aaf..000000000000
--- a/arch/x86/um/stub_32.S
+++ /dev/null
@@ -1,56 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0 */
-#include <as-layout.h>
-
-.section .__syscall_stub, "ax"
-
-	.globl batch_syscall_stub
-batch_syscall_stub:
-	/* %esp comes in as "top of page" */
-	mov %esp, %ecx
-	/* %esp has pointer to first operation */
-	add $8, %esp
-again:
-	/* load length of additional data */
-	mov	0x0(%esp), %eax
-
-	/* if(length == 0) : end of list */
-	/* write possible 0 to header */
-	mov	%eax, 0x4(%ecx)
-	cmpl	$0, %eax
-	jz	done
-
-	/* save current pointer */
-	mov	%esp, 0x4(%ecx)
-
-	/* skip additional data */
-	add	%eax, %esp
-
-	/* load syscall-# */
-	pop	%eax
-
-	/* load syscall params */
-	pop	%ebx
-	pop	%ecx
-	pop	%edx
-	pop	%esi
- 	pop	%edi
-	pop	%ebp
-
-	/* execute syscall */
-	int	$0x80
-
-	/* restore top of page pointer in %ecx */
-	mov	%esp, %ecx
-	andl	$(~UM_KERN_PAGE_SIZE) + 1, %ecx
-
-	/* check return value */
-	pop	%ebx
-	cmp	%ebx, %eax
-	je	again
-
-done:
-	/* save return value */
-	mov	%eax, (%ecx)
-
-	/* stop */
-	int3
diff --git a/arch/x86/um/stub_64.S b/arch/x86/um/stub_64.S
deleted file mode 100644
index f3404640197a..000000000000
--- a/arch/x86/um/stub_64.S
+++ /dev/null
@@ -1,50 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0 */
-#include <as-layout.h>
-
-.section .__syscall_stub, "ax"
-	.globl batch_syscall_stub
-batch_syscall_stub:
-	/* %rsp has the pointer to first operation */
-	mov	%rsp, %rbx
-	add	$0x10, %rsp
-again:
-	/* load length of additional data */
-	mov	0x0(%rsp), %rax
-
-	/* if(length == 0) : end of list */
-	/* write possible 0 to header */
-	mov	%rax, 8(%rbx)
-	cmp	$0, %rax
-	jz	done
-
-	/* save current pointer */
-	mov	%rsp, 8(%rbx)
-
-	/* skip additional data */
-	add	%rax, %rsp
-
-	/* load syscall-# */
-	pop	%rax
-
-	/* load syscall params */
-	pop	%rdi
-	pop	%rsi
-	pop	%rdx
-	pop	%r10
- 	pop	%r8
-	pop	%r9
-
-	/* execute syscall */
-	syscall
-
-	/* check return value */
-	pop	%rcx
-	cmp	%rcx, %rax
-	je	again
-
-done:
-	/* save return value */
-	mov	%rax, (%rbx)
-
-	/* stop */
-	int3