Re: SAR implementation in Bochs
"Stanislav" <[email protected]> Thu, 18 Dec 2014 21:35:21 +0200
| Newsgroups | gmane.comp.emulators.bochs.devel |
|---|---|
| Message-ID | <[email protected]> |
I feel safe with the current code. Yes, in theory it is implementation dependent but in practice I've never seen any implementation which does it wrong way and not doing an arithmetic shift right when the left operand is signed. Bochs has current code for years and was compiled probably for every possible platform using various compilers and nobody complained till now. Doing it "safe" way would be also much slower. Stanislav From: [email protected] [mailto:[email protected]] Sent: Thursday, December 18, 2014 6:09 PM To: [email protected] Subject: SAR implementation in Bochs Hello, It seems that the SAR instruction implementation in Bochs is kind of dangerous, as doing a bitshift on a negative value is implementation dependant, it might not end up the same on all platforms. I think the worst thing that can happen is an off-by-one error in the result, but that is enough to break programs... Please correct me if I'm wrong! * Olav _____ This message was sent to you via the SourceForge web mail form. You may reply to this message directly, or at http://sourceforge.net/u/eightbitbubsy/profile/send_message ------------------------------------------------------------------------------ Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT Server from Actuate! Instantly Supercharge Your Business Reports and Dashboards with Interactivity, Sharing, Native Excel Exports, App Integration & more Get technology previously reserved for billion-dollar corporations, FREE http://pubads.g.doubleclick.net/gampad/clk?id=164703151&iu=/4140/ostg.clktrk _______________________________________________ bochs-developers mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/bochs-developers