unwind: Fix unwinder for when last instruction is a call

When a function's last instruction is a BL (Branch and Link), the
address of the next instruction (outside this function) is kept in the
LR. The callee function then saves this LR on stack as part of the frame
record.

When unwinding the stack from this callee function, the unwinder
recovers this LR from the frame record and uses this address to find out
the caller. As this address is beyond the caller, the unwinder fails to
find it in the sframe table for any function and fails.

Fix this edge case by assuming 1 extra instruction after the function
as part of the function when searching the sframe table.

NOTE: This could have a side affect that when a frame record is
malformed/corrupted, the LR can point to the end of a function + one
instruction, even without that function's last instruction being a
branch. But when the frame record is corrupted at runtime, there is no
way to figure that out anyway so this change doesn't add to or establish
a new problem.

devtmpfsd() is a function that displays this behaviour in the kernel:
-------------------------------------------------------------------------------------
ffff800080d9a070 <devtmpfsd>:
ffff800080d9a070:       d503201f        nop
ffff800080d9a074:       d503201f        nop
ffff800080d9a078:       d503233f        paciasp
ffff800080d9a07c:       a9be7bfd        stp     x29, x30, [sp, #-32]!
ffff800080d9a080:       910003fd        mov     x29, sp
ffff800080d9a084:       f9000bf3        str     x19, [sp, gregkh#16]
[...SNIP....]
ffff800080d9a0a0:       340000d3        cbz     w19, ffff800080d9a0b8
[....SNIP...]
ffff800080d9a0ac:       a8c27bfd        ldp     x29, x30, [sp], #32
ffff800080d9a0b0:       d50323bf        autiasp
ffff800080d9a0b4:       d65f03c0        ret
ffff800080d9a0b8:       97f06526        bl      ffff8000809b3550 <devtmpfs_work_loop>
-------------------------------------------------------------------------------------
ffff800080d9a0bc:       00000000        udf     #0

Here devtmpfsd() calls devtmpfs_work_loop() and the compiler know the
devtmpfs_work_loop() never returns, so it adds the BL to this function
as the last instruction of devtmpfsd()

Signed-off-by: Puranjay Mohan <pjy@amazon.com>

Author: puranjaymohan

kernel/sframe_lookup.c

@@ -83,7 +83,7 @@ static struct sframe_fde *find_fde(const struct sframe_table *tbl, unsigned long
 	if (f >= tbl->sfhdr_p->num_fdes || f < 0)
 		return NULL;
 	fdep = tbl->fde_p + f;
-	if (ip < fdep->start_addr || ip >= fdep->start_addr + fdep->size)
+	if (ip < fdep->start_addr || ip > fdep->start_addr + fdep->size)
 		return NULL;
 
 	return fdep;
@@ -107,7 +107,7 @@ static int find_fre(const struct sframe_table *tbl, unsigned long pc,
 	else
 		ip_off = (int32_t)(pc - (unsigned long)tbl->sfhdr_p) - fdep->start_addr;
 
-	if (ip_off < 0 || ip_off >= fdep->size)
+	if (ip_off < 0 || ip_off > fdep->size)
 		return -EINVAL;
 
 	/*