<div dir="ltr">Hi Nicolas,<div><br></div><div> good catch. thanks.<br><div class="gmail_extra"><br><br><div class="gmail_quote">On Mon, Dec 16, 2013 at 2:36 PM, Nicolas Cellier <span dir="ltr"><<a href="mailto:nicolas.cellier.aka.nice@gmail.com" target="_blank">nicolas.cellier.aka.nice@gmail.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> <br><div dir="ltr"><div>(SmallInteger new: 0) can quickly lead to a VM crash.<br></div><div>This should better be intercepted as a primitive failure.<br>
</div></div></blockquote><div><br></div><div>greed. But there are two obvious ways to go about this (at least in Spur). One is to explicitly test for the immediate classes in the new and new: primitives. This is poor as it introduces additional tests on a critical path. The other way is to give SmallInteger and Character special formats and then the primitives would fail without additional tests. The question is what format to use. In Spur the formats are</div>
<div><br></div><div><div><span class="Apple-tab-span" style="white-space:pre"> </span> 0 = 0 sized objects (UndefinedObject True False et al)</div><div><span class="Apple-tab-span" style="white-space:pre"> </span> 1 = non-indexable objects with inst vars (Point et al)</div>
<div><span class="Apple-tab-span" style="white-space:pre"> </span> 2 = indexable objects with no inst vars (Array et al)</div><div><span class="Apple-tab-span" style="white-space:pre"> </span> 3 = indexable objects with inst vars (MethodContext AdditionalMethodState et al)</div>
<div><span class="Apple-tab-span" style="white-space:pre"> </span> 4 = weak indexable objects with inst vars (WeakArray et al)</div><div><span class="Apple-tab-span" style="white-space:pre"> </span> 5 = weak non-indexable objects with inst vars (ephemerons) (Ephemeron)</div>
<div><span class="Apple-tab-span" style="white-space:pre"> </span> 6 unused, reserved for exotic pointer objects?</div><div><span class="Apple-tab-span" style="white-space:pre"> </span> 7 Forwarded Object, 1st field is pointer, rest of fields are ignored</div>
<div><span class="Apple-tab-span" style="white-space:pre"> </span> 8 unused, reserved for exotic non-pointer objects?</div><div><span class="Apple-tab-span" style="white-space:pre"> </span> 9 (?) 64-bit indexable</div><div>
<span class="Apple-tab-span" style="white-space:pre"> </span> 10 - 11 32-bit indexable<span class="Apple-tab-span" style="white-space:pre"> </span>(11 unused in 32 bits)</div><div><span class="Apple-tab-span" style="white-space:pre"> </span> 12 - 15 16-bit indexable<span class="Apple-tab-span" style="white-space:pre"> </span>(14 & 15 unused in 32-bits)</div>
<div><span class="Apple-tab-span" style="white-space:pre"> </span> 16 - 23 byte indexable<span class="Apple-tab-span" style="white-space:pre"> </span>(20-23 unused in 32-bits)</div><div><span class="Apple-tab-span" style="white-space:pre"> </span> 24 - 31 compiled method<span class="Apple-tab-span" style="white-space:pre"> </span>(28-21 unused in 32-bits)</div>
</div><div><br></div><div>Format 7 should never be valid for an instance as it is reserved by the VM for forwarders. So how about making 7 the format for immediate classes? Ten it will be easy to arrange that both new and new: will fail naturally for this format.</div>
<div><br></div><div><br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr"><div></div>I know that we cannot protect against all and every mistake, but this one should be easy.<br>
<div><div><br>See <a href="http://stackoverflow.com/questions/20621119/how-to-print-an-integer-to-transcript" target="_blank">http://stackoverflow.com/questions/20621119/how-to-print-an-integer-to-transcript</a><br><br></div>
<div>I'm thinking of Spur where Character are immediate and where archeo-code could lead to same kind of crash.</div></div></div></blockquote></div><div><br></div>-- <br>best,<div>Eliot</div>
</div></div></div>