[llvm-bugs] [Bug 52036] New: Clang breaks badly on too large object sizes

via llvm-bugs llvm-bugs at lists.llvm.org
Fri Oct 1 23:18:55 PDT 2021


            Bug ID: 52036
           Summary: Clang breaks badly on too large object sizes
           Product: clang
           Version: unspecified
          Hardware: PC
                OS: Linux
            Status: NEW
          Severity: enhancement
          Priority: P
         Component: C
          Assignee: unassignedclangbugs at nondot.org
          Reporter: gabravier at gmail.com
                CC: blitzrakete at gmail.com, dgregor at apple.com,
                    erik.pilkington at gmail.com, llvm-bugs at lists.llvm.org,
                    richard-llvm at metafoo.co.uk

struct s {
    char arr[922337203685477587];
    char arr2[922337203685477587];
    char arr3[922337203685477587];
    char arr4[922337203685477587];
    char arr5[922337203685477587];
    char arr6[922337203685477587];
    char arr7[922337203685477587];
    char arr8[922337203685477587];
    char arr9[922337203685477587];
    char arr10[922337203685477587];
    char arr11[922337203685477587];
    char arr12[922337203685477587];

s x;

void f()
    volatile s x = {};

The size of `s` is simply not representable in `size_t`, which seems to break
Clang in multiple ways:
- It gives me a warning telling me that "stack frame size exceeds limit" when
the warning tells me of a stack frame size that's smaller than the limit (it
tells me the stack frame size is 1717986984 and that the limit is 4294967295)
- The emitted assembly tries to call memset with a size of 461168601842738840,
which is smaller than even a single one of the arrays in `s`.

Probably Clang should just error on the declaration of `s`, as GCC does:

<source>:1:8: error: size of type 's' is too large ('11068046444225731044'
    1 | struct s {
      |        ^

You are receiving this mail because:
You are on the CC list for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-bugs/attachments/20211002/ef4458d4/attachment.html>

More information about the llvm-bugs mailing list