Thread (53 messages) 53 messages, 4 authors, 2020-05-05

Re: [PATCH v7 02/28] powerpc/xmon: Move breakpoint instructions to own array

From: Jordan Niethe <hidden>
Date: 2020-05-04 05:55:09

On Mon, May 4, 2020 at 3:41 PM Alistair Popple [off-list ref] wrote:
On Friday, 1 May 2020 1:41:54 PM AEST Jordan Niethe wrote:
quoted
To execute an instruction out of line after a breakpoint, the NIP is set
to the address of struct bpt::instr. Here a copy of the instruction that
was replaced with a breakpoint is kept, along with a trap so normal flow
can be resumed after XOLing. The struct bpt's are located within the
data section. This is problematic as the data section may be marked as
no execute.

Instead of each struct bpt holding the instructions to be XOL'd, make a
new array, bpt_table[], with enough space to hold instructions for the
number of supported breakpoints. A later patch will move this to the
text section.
Make struct bpt::instr a pointer to the instructions in bpt_table[]
associated with that breakpoint. This association is a simple mapping:
bpts[n] -> bpt_table[n * words per breakpoint]. Currently we only need
the copied instruction followed by a trap, so 2 words per breakpoint.

Signed-off-by: Jordan Niethe <redacted>
---
v4: New to series
v5: - Do not use __section(), use a .space directive in .S file
    - Simplify in_breakpoint_table() calculation
    - Define BPT_SIZE
v6: - Seperate moving to text section
---
 arch/powerpc/xmon/xmon.c | 21 ++++++++++++---------
 1 file changed, 12 insertions(+), 9 deletions(-)
diff --git a/arch/powerpc/xmon/xmon.c b/arch/powerpc/xmon/xmon.c
index f91ae2c9adbe..6ba7f66c1dd0 100644
--- a/arch/powerpc/xmon/xmon.c
+++ b/arch/powerpc/xmon/xmon.c
@@ -98,7 +98,7 @@ static long *xmon_fault_jmp[NR_CPUS];
 /* Breakpoint stuff */
 struct bpt {
      unsigned long   address;
-     unsigned int    instr[2];
+     unsigned int    *instr;
      atomic_t        ref_count;
      int             enabled;
      unsigned long   pad;
@@ -117,6 +117,10 @@ static unsigned bpinstr = 0x7fe00008;    /* trap */

 #define BP_NUM(bp)   ((bp) - bpts + 1)

+#define BPT_SIZE       (sizeof(unsigned int) * 2)
+#define BPT_WORDS      (BPT_SIZE / sizeof(unsigned int))
Minor nit-pick but IMHO this would be more logical if you defined BPT_WORDS
first like so:

#define BPT_WORDS      (2)
#define BPT_SIZE       (sizeof(unsigned int) * BPT_WORDS)

Otherwise this looks good and I think the offset calculations below are correct
so:
What I was thinking was BPT_SIZE  would later be defined in terms the
instruction type and by doing it this way
 BPT_WORDS  would be correct for the the instruction type if like on
ppc32, it did not include a suffix.
Reviewed-by: Alistair Popple <redacted>
quoted
+static unsigned int bpt_table[NBPTS * BPT_WORDS];
+
 /* Prototypes */
 static int cmds(struct pt_regs *);
 static int mread(unsigned long, void *, int);
@@ -854,15 +858,13 @@ static struct bpt *in_breakpoint_table(unsigned long
nip, unsigned long *offp) {
      unsigned long off;

-     off = nip - (unsigned long) bpts;
-     if (off >= sizeof(bpts))
+     off = nip - (unsigned long) bpt_table;
+     if (off >= sizeof(bpt_table))
              return NULL;
-     off %= sizeof(struct bpt);
-     if (off != offsetof(struct bpt, instr[0])
-         && off != offsetof(struct bpt, instr[1]))
+     *offp = off % BPT_SIZE;
+     if (*offp != 0 && *offp != 4)
              return NULL;
-     *offp = off - offsetof(struct bpt, instr[0]);
-     return (struct bpt *) (nip - off);
+     return bpts + (off / BPT_SIZE);
 }

 static struct bpt *new_breakpoint(unsigned long a)
@@ -877,7 +879,8 @@ static struct bpt *new_breakpoint(unsigned long a)
      for (bp = bpts; bp < &bpts[NBPTS]; ++bp) {
              if (!bp->enabled && atomic_read(&bp->ref_count) == 0) {
                      bp->address = a;
-                     patch_instruction(&bp->instr[1], bpinstr);
+                     bp->instr = bpt_table + ((bp - bpts) * BPT_WORDS);
+                     patch_instruction(bp->instr + 1, bpinstr);
                      return bp;
              }
      }

Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help