Re: [fpc-devel] Bitpacked Record Bug

2010-10-26 Thread Jonas Maebe
On 24 Oct 2010, at 23:56, Willibald Krenn wrote: Am 24.10.2010 20:49, schrieb Jonas Maebe: It should probably check whether the packedbitsize is in [8, 16, 32, 64]. The reason is that if you have a load from a byte-aligned address using a size that is natively supported by the code

Re: [fpc-devel] Bitpacked Record Bug

2010-10-26 Thread Willibald Krenn
Am 26.10.2010 21:33, schrieb Jonas Maebe: Thanks for confirming my thoughts. Shall I still file a bugreport for this? It's fixed. Thanks, I just wanted to file a report for this and the assembly thing but it seems that reporting new issues on mantis does not work currently. (I am getting

Re: [fpc-devel] Bitpacked Record Bug

2010-10-26 Thread Sergei Gorelkin
Willibald Krenn wrote: Am 26.10.2010 21:33, schrieb Jonas Maebe: Thanks for confirming my thoughts. Shall I still file a bugreport for this? It's fixed. Thanks, I just wanted to file a report for this and the assembly thing but it seems that reporting new issues on mantis does not work

Re: [fpc-devel] Bitpacked Record Bug

2010-10-24 Thread Florian Klämpfl
Am 23.10.2010 15:47, schrieb Willibald Krenn: Hi! I stumbled accross a bug in fpc that manifests itself when working with bitpacked records that have fields with pos and size mod 8 == 0; For example: type TField = bitpacked record case boolean of true: (value : integer);

Re: [fpc-devel] Bitpacked Record Bug

2010-10-24 Thread Jonas Maebe
On 23 Oct 2010, at 15:47, Willibald Krenn wrote: in nutils.pas (~ 1155) subscriptn: result:= is_packed_record_or_object(tsubscriptnode(n).left.resultdef) and ( (tsubscriptnode(n).vs.vardef.packedbitsize mod 8 0) or (tsubscriptnode(n).vs.fieldoffset mod 8 0));

Re: [fpc-devel] Bitpacked Record Bug

2010-10-24 Thread Willibald Krenn
Am 24.10.2010 20:49, schrieb Jonas Maebe: It should probably check whether the packedbitsize is in [8, 16, 32, 64]. The reason is that if you have a load from a byte-aligned address using a size that is natively supported by the code generator, you don't have to go through all the

Re: [fpc-devel] Bitpacked Record Bug

2010-10-24 Thread Jonas Maebe
On 24 Oct 2010, at 23:56, Willibald Krenn wrote: Am 24.10.2010 20:49, schrieb Jonas Maebe: It should probably check whether the packedbitsize is in [8, 16, 32, 64]. The reason is that if you have a load from a byte-aligned address using a size that is natively supported by the code

[fpc-devel] Bitpacked Record Bug

2010-10-23 Thread Willibald Krenn
Hi! I stumbled accross a bug in fpc that manifests itself when working with bitpacked records that have fields with pos and size mod 8 == 0; For example: type TField = bitpacked record case boolean of true: (value : integer); false: ( data: 0 .. $FF;

Re: [fpc-devel] Bitpacked Record Bug

2010-10-23 Thread Willibald Krenn
Am 23.10.2010 14:47, schrieb Willibald Krenn: BTW: Changing is_packed_record to in nutils.pas (~ 1155) subscriptn: result:= is_packed_record_or_object(tsubscriptnode(n).left.resultdef) and ( (tsubscriptnode(n).vs.vardef.packedbitsize mod 8 0) or (tsubscriptnode(n).vs.fieldoffset mod 8 0));