qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Paolo Bonzini <pbonzini@redhat.com>
To: qemu-devel@nongnu.org
Cc: "Philippe Mathieu-Daudé" <philmd@redhat.com>
Subject: [PULL v2 25/31] osdep: Make MIN/MAX evaluate arguments only once
Date: Fri, 26 Jun 2020 09:56:08 -0400	[thread overview]
Message-ID: <20200626135608.6920-3-pbonzini@redhat.com> (raw)
In-Reply-To: <20200626135608.6920-1-pbonzini@redhat.com>

From: Eric Blake <eblake@redhat.com>

I'm not aware of any immediate bugs in qemu where a second runtime
evaluation of the arguments to MIN() or MAX() causes a problem, but
proactively preventing such abuse is easier than falling prey to an
unintended case down the road.  At any rate, here's the conversation
that sparked the current patch:
https://lists.gnu.org/archive/html/qemu-devel/2018-12/msg05718.html

Update the MIN/MAX macros to only evaluate their argument once at
runtime; this uses typeof(1 ? (a) : (b)) to ensure that we are
promoting the temporaries to the same type as the final comparison (we
have to trigger type promotion, as typeof(bitfield) won't compile; and
we can't use typeof((a) + (b)) or even typeof((a) + 0), as some of our
uses of MAX are on void* pointers where such addition is undefined).

However, we are unable to work around gcc refusing to compile ({}) in
a constant context (such as the array length of a static variable),
even when only used in the dead branch of a __builtin_choose_expr(),
so we have to provide a second macro pair MIN_CONST and MAX_CONST for
use when both arguments are known to be compile-time constants and
where the result must also be usable as a constant; this second form
evaluates arguments multiple times but that doesn't matter for
constants.  By using a void expression as the expansion if a
non-constant is presented to this second form, we can enlist the
compiler to ensure the double evaluation is not attempted on
non-constants.

Alas, as both macros now rely on compiler intrinsics, they are no
longer usable in preprocessor #if conditions; those will just have to
be open-coded or the logic rewritten into #define or runtime 'if'
conditions (but where the compiler dead-code-elimination will probably
still apply).

I tested that both gcc 10.1.1 and clang 10.0.0 produce errors for all
forms of macro mis-use.  As the errors can sometimes be cryptic, I'm
demonstrating the gcc output:

Use of MIN when MIN_CONST is needed:

In file included from /home/eblake/qemu/qemu-img.c:25:
/home/eblake/qemu/include/qemu/osdep.h:249:5: error: braced-group within expression allowed only inside a function
  249 |     ({                                                  \
      |     ^
/home/eblake/qemu/qemu-img.c:92:12: note: in expansion of macro ‘MIN’
   92 | char array[MIN(1, 2)] = "";
      |            ^~~

Use of MIN_CONST when MIN is needed:

/home/eblake/qemu/qemu-img.c: In function ‘is_allocated_sectors’:
/home/eblake/qemu/qemu-img.c:1225:15: error: void value not ignored as it ought to be
 1225 |             i = MIN_CONST(i, n);
      |               ^

Use of MIN in the preprocessor:

In file included from /home/eblake/qemu/accel/tcg/translate-all.c:20:
/home/eblake/qemu/accel/tcg/translate-all.c: In function ‘page_check_range’:
/home/eblake/qemu/include/qemu/osdep.h:249:6: error: token "{" is not valid in preprocessor expressions
  249 |     ({                                                  \
      |      ^

Fix the resulting callsites that used #if or computed a compile-time
constant min or max to use the new macros.  cpu-defs.h is interesting,
as CPU_TLB_DYN_MAX_BITS is sometimes used as a constant and sometimes
dynamic.

It may be worth improving glib's MIN/MAX definitions to be saner, but
that is a task for another day.

Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20200625162602.700741-1-eblake@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
 accel/tcg/translate-all.c |  6 ++---
 hw/usb/hcd-xhci.h         |  2 +-
 include/block/block.h     |  4 +--
 include/exec/cpu-all.h    |  8 +++---
 include/exec/cpu-defs.h   |  7 ++++-
 include/qemu/osdep.h      | 57 ++++++++++++++++++++++++++++++++-------
 migration/qemu-file.c     |  2 +-
 7 files changed, 63 insertions(+), 23 deletions(-)

diff --git a/accel/tcg/translate-all.c b/accel/tcg/translate-all.c
index c3d37058a1..2afa46bd2b 100644
--- a/accel/tcg/translate-all.c
+++ b/accel/tcg/translate-all.c
@@ -2582,9 +2582,9 @@ int page_check_range(target_ulong start, target_ulong len, int flags)
     /* This function should never be called with addresses outside the
        guest address space.  If this assert fires, it probably indicates
        a missing call to h2g_valid.  */
-#if TARGET_ABI_BITS > L1_MAP_ADDR_SPACE_BITS
-    assert(start < ((target_ulong)1 << L1_MAP_ADDR_SPACE_BITS));
-#endif
+    if (TARGET_ABI_BITS > L1_MAP_ADDR_SPACE_BITS) {
+        assert(start < ((target_ulong)1 << L1_MAP_ADDR_SPACE_BITS));
+    }
 
     if (len == 0) {
         return 0;
diff --git a/hw/usb/hcd-xhci.h b/hw/usb/hcd-xhci.h
index 2fad4df2a7..946af51fc2 100644
--- a/hw/usb/hcd-xhci.h
+++ b/hw/usb/hcd-xhci.h
@@ -214,7 +214,7 @@ struct XHCIState {
     uint32_t dcbaap_high;
     uint32_t config;
 
-    USBPort  uports[MAX(MAXPORTS_2, MAXPORTS_3)];
+    USBPort  uports[MAX_CONST(MAXPORTS_2, MAXPORTS_3)];
     XHCIPort ports[MAXPORTS];
     XHCISlot slots[MAXSLOTS];
     uint32_t numports;
diff --git a/include/block/block.h b/include/block/block.h
index 25e299605e..e8fc814996 100644
--- a/include/block/block.h
+++ b/include/block/block.h
@@ -133,8 +133,8 @@ typedef struct HDGeometry {
 #define BDRV_SECTOR_BITS   9
 #define BDRV_SECTOR_SIZE   (1ULL << BDRV_SECTOR_BITS)
 
-#define BDRV_REQUEST_MAX_SECTORS MIN(SIZE_MAX >> BDRV_SECTOR_BITS, \
-                                     INT_MAX >> BDRV_SECTOR_BITS)
+#define BDRV_REQUEST_MAX_SECTORS MIN_CONST(SIZE_MAX >> BDRV_SECTOR_BITS, \
+                                           INT_MAX >> BDRV_SECTOR_BITS)
 #define BDRV_REQUEST_MAX_BYTES (BDRV_REQUEST_MAX_SECTORS << BDRV_SECTOR_BITS)
 
 /*
diff --git a/include/exec/cpu-all.h b/include/exec/cpu-all.h
index fb4e8a8e29..fc403d456b 100644
--- a/include/exec/cpu-all.h
+++ b/include/exec/cpu-all.h
@@ -176,11 +176,9 @@ extern unsigned long reserved_va;
  * avoid setting bits at the top of guest addresses that might need
  * to be used for tags.
  */
-#if MIN(TARGET_VIRT_ADDR_SPACE_BITS, TARGET_ABI_BITS) <= 32
-# define GUEST_ADDR_MAX_  UINT32_MAX
-#else
-# define GUEST_ADDR_MAX_  (~0ul)
-#endif
+#define GUEST_ADDR_MAX_                                                 \
+    ((MIN_CONST(TARGET_VIRT_ADDR_SPACE_BITS, TARGET_ABI_BITS) <= 32) ?  \
+     UINT32_MAX : ~0ul)
 #define GUEST_ADDR_MAX    (reserved_va ? reserved_va - 1 : GUEST_ADDR_MAX_)
 
 #else
diff --git a/include/exec/cpu-defs.h b/include/exec/cpu-defs.h
index 8c44abefa2..9185632337 100644
--- a/include/exec/cpu-defs.h
+++ b/include/exec/cpu-defs.h
@@ -102,8 +102,13 @@ typedef uint64_t target_ulong;
  * Skylake's Level-2 STLB has 16 1G entries.
  * Also, make sure we do not size the TLB past the guest's address space.
  */
-#  define CPU_TLB_DYN_MAX_BITS                                  \
+#  ifdef TARGET_PAGE_BITS_VARY
+#   define CPU_TLB_DYN_MAX_BITS                                  \
     MIN(22, TARGET_VIRT_ADDR_SPACE_BITS - TARGET_PAGE_BITS)
+#  else
+#   define CPU_TLB_DYN_MAX_BITS                                  \
+    MIN_CONST(22, TARGET_VIRT_ADDR_SPACE_BITS - TARGET_PAGE_BITS)
+#  endif
 # endif
 
 typedef struct CPUTLBEntry {
diff --git a/include/qemu/osdep.h b/include/qemu/osdep.h
index ff7c17b857..0d26a1b9bd 100644
--- a/include/qemu/osdep.h
+++ b/include/qemu/osdep.h
@@ -236,18 +236,55 @@ extern int daemon(int, int);
 #define SIZE_MAX ((size_t)-1)
 #endif
 
-#ifndef MIN
-#define MIN(a, b) (((a) < (b)) ? (a) : (b))
-#endif
-#ifndef MAX
-#define MAX(a, b) (((a) > (b)) ? (a) : (b))
-#endif
+/*
+ * Two variations of MIN/MAX macros. The first is for runtime use, and
+ * evaluates arguments only once (so it is safe even with side
+ * effects), but will not work in constant contexts (such as array
+ * size declarations) because of the '{}'.  The second is for constant
+ * expression use, where evaluating arguments twice is safe because
+ * the result is going to be constant anyway, but will not work in a
+ * runtime context because of a void expression where a value is
+ * expected.  Thus, both gcc and clang will fail to compile if you use
+ * the wrong macro (even if the error may seem a bit cryptic).
+ *
+ * Note that neither form is usable as an #if condition; if you truly
+ * need to write conditional code that depends on a minimum or maximum
+ * determined by the pre-processor instead of the compiler, you'll
+ * have to open-code it.
+ */
+#undef MIN
+#define MIN(a, b)                                       \
+    ({                                                  \
+        typeof(1 ? (a) : (b)) _a = (a), _b = (b);       \
+        _a < _b ? _a : _b;                              \
+    })
+#define MIN_CONST(a, b)                                         \
+    __builtin_choose_expr(                                      \
+        __builtin_constant_p(a) && __builtin_constant_p(b),     \
+        (a) < (b) ? (a) : (b),                                  \
+        ((void)0))
+#undef MAX
+#define MAX(a, b)                                       \
+    ({                                                  \
+        typeof(1 ? (a) : (b)) _a = (a), _b = (b);       \
+        _a > _b ? _a : _b;                              \
+    })
+#define MAX_CONST(a, b)                                         \
+    __builtin_choose_expr(                                      \
+        __builtin_constant_p(a) && __builtin_constant_p(b),     \
+        (a) > (b) ? (a) : (b),                                  \
+        ((void)0))
 
-/* Minimum function that returns zero only iff both values are zero.
- * Intended for use with unsigned values only. */
+/*
+ * Minimum function that returns zero only if both values are zero.
+ * Intended for use with unsigned values only.
+ */
 #ifndef MIN_NON_ZERO
-#define MIN_NON_ZERO(a, b) ((a) == 0 ? (b) : \
-                                ((b) == 0 ? (a) : (MIN(a, b))))
+#define MIN_NON_ZERO(a, b)                              \
+    ({                                                  \
+        typeof(1 ? (a) : (b)) _a = (a), _b = (b);       \
+        _a == 0 ? _b : (_b == 0 || _b > _a) ? _a : _b;  \
+    })
 #endif
 
 /* Round number down to multiple */
diff --git a/migration/qemu-file.c b/migration/qemu-file.c
index 1c3a358a14..be21518c57 100644
--- a/migration/qemu-file.c
+++ b/migration/qemu-file.c
@@ -31,7 +31,7 @@
 #include "qapi/error.h"
 
 #define IO_BUF_SIZE 32768
-#define MAX_IOV_SIZE MIN(IOV_MAX, 64)
+#define MAX_IOV_SIZE MIN_CONST(IOV_MAX, 64)
 
 struct QEMUFile {
     const QEMUFileOps *ops;
-- 
2.26.2




  parent reply	other threads:[~2020-06-26 13:57 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-06-26 13:56 [PULL v2 00/31] Misc patches for 2020-06-24 Paolo Bonzini
2020-06-26 13:56 ` [PULL v2 09/31] exec: fetch the alignment of Linux devdax pmem character device nodes Paolo Bonzini
2020-06-26 13:56 ` Paolo Bonzini [this message]
2020-06-27 21:35   ` [PULL v2 25/31] osdep: Make MIN/MAX evaluate arguments only once Peter Maydell
2020-06-29 15:23     ` Eric Blake
2020-06-29 15:28       ` Paolo Bonzini
2020-06-29 16:01     ` Markus Armbruster
2020-06-26 17:22 ` [PULL v2 00/31] Misc patches for 2020-06-24 Peter Maydell

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20200626135608.6920-3-pbonzini@redhat.com \
    --to=pbonzini@redhat.com \
    --cc=philmd@redhat.com \
    --cc=qemu-devel@nongnu.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).