similar to: [LLVMdev] Disabling RIP-Relative Addressing

On May 16, 2011, at 6:24 AM, Chris Stavrakakis wrote: > Hi all, > > in llvm-2.8, i declare an external constant of i64 and i want the > instructions that use the value of this constant to use absolute > addressing instead of RIP relative, that is: > > I get: "48 8b 35 00 00 00 00 mov 0x0(%rip),%rsi" #AT&T syntax > But i would prefer: "48 BE

Hi all, in llvm-2.8, i declare an external constant of i64 and i want the instructions that use the value of this constant to use absolute addressing instead of RIP relative, that is: I get: "48 8b 35 00 00 00 00 mov 0x0(%rip),%rsi" #AT&T syntax But i would prefer: "48 BE 00 00 00 00 00 00 mov 0x0, %rsi" The reason i want this is that i use a specific

Chris, > You'll have to hack up the code generator, there is no user visible knob to turn off rip relative addressing. Why? One can use medium code model which does not do any rip-rel stuff. -- With best regards, Anton Korobeynikov Faculty of Mathematics and Mechanics, Saint Petersburg State University

On 05/17/2011 07:08 PM, Chris Lattner wrote: > Oh right, that could work. > > -Chris > > On May 17, 2011, at 6:25 AM, Anton Korobeynikov<anton at korobeynikov.info> wrote: > >> Chris, >> >>> You'll have to hack up the code generator, there is no user visible knob to turn off rip relative addressing. >> Why? One can use medium code model which

Apologies - I apparently remembered part of the issue incorrectly, so this ended up quite confusing. The problem comes when referencing labels in a different section of the binary. To clarify, if I assemble the code: .data foo BYTE 5 .code mov eax, foo with Microsoft's ml64.exe, it emits an object file disassembling to: 0: 8b 05 00 00 00 00 mov eax, dword ptr [rip]

Are you asking what the parsing rules are, or how you should modify the LLVM code to achieve that result? If the latter, you haven’t really given enough detail here. What code, exactly, have you tried modifying? Do you have any ideas for how it could work? -Eli From: Eric Astor <epastor at google.com> Sent: Tuesday, January 21, 2020 2:44 PM To: Eli Friedman <efriedma at

Hi all, Continuing work on llvm-ml (a MASM assembler)... and my latest obstacle is in enabling MASM's convention that (unless specified) all memory location references should be RIP-relative. Without it, we emit the wrong instructions for "call", "jmp", etc., and anything we build fails at the linking stage. My best attempt at this so far is a small patch to

Author: dmick Repository: /hg/zfs-crypto/gate Revision: 2120ccf2018170cfe16915dac09370ad30dc5285 Log message: 6251453 dis should decode rip-relative memory accesses 6279427 mdb''s x64 disassembler doesn''t decode %rip-relative addresses for data access 6427698 mdb/kmdb/dis should look up symbols for immediate operands 6428349 mdb/kmdb/dis (libdisasm) show odd offset for x86

This saves one insn byte per instance, summing up to a savings of over 4k in my (stripped down) configuration. No variant of to be patched in replacement code relies on the one byte larger size. Signed-off-by: Jan Beulich <jbeulich at suse.com> --- arch/x86/include/asm/paravirt_types.h | 6 ++++++ 1 file changed, 6 insertions(+) --- 4.18-rc2/arch/x86/include/asm/paravirt_types.h +++

The destination register and each of the source registers can be relatively addressed by some special "address registers": a0, a1, a2, a3. suppose a0 = 4, a1 = 2, the instruction add r[a0], r[a1], r7 equals to: r4 = r2 + r7 How should the instruction be defined in TableGen *.td file? If an opcode is defined for each variant (like X86InstrInfo.td) there will be eight opcodes

On 25/06/18 12:29, Jan Beulich wrote: > This saves one insn byte per instance, summing up to a savings of over > 4k in my (stripped down) configuration. No variant of to be patched in > replacement code relies on the one byte larger size. > > Signed-off-by: Jan Beulich <jbeulich at suse.com> Reviewed-by: Juergen Gross <jgross at suse.com> Juergen

On Mon, May 23, 2016 at 2:44 PM, Jun Koi <junkoi2004 at gmail.com> wrote: > Hi, > > I found a mismatch between assembler & disassembler of X86: I assembled an > instruction, then disassembled the output, but the result is not the same > as the first original code: "add qword ptr [205163983024656], 1" vs " add > qword ptr [1985229328], 1" > >

Hi, I found a mismatch between assembler & disassembler of X86: I assembled an instruction, then disassembled the output, but the result is not the same as the first original code: "add qword ptr [205163983024656], 1" vs " add qword ptr [1985229328], 1" Anybody knows what is wrong? Thanks. $ echo "ADD QWORD PTR [0xba9876543210], 0x1"|llvm-mc -assemble

+1 IPv6 = solution looking for a problem. Disabled on all our systems! -----Original Message----- From: centos-bounces at centos.org [mailto:centos-bounces at centos.org] On Behalf Of Chris Stone Sent: Monday, March 09, 2015 01:15 AM To: CentOS mailing list Subject: Re: [CentOS] Centos 6 - disabling IPv6 addressing Sorry - that should be sysctl -w net.ipv6.conf.all.accept_ra=0 to disable

Hi all, I have installed a CentOS-8 server to accomplish some tests using kvm guests ? All these guests are centos8 also. When these guests starts, I always see errors like: [ 589.687915] kvm [2133]: vcpu0, guest rIP: 0xffffffff81a5bc98 disabled perfctr wrmsr: 0xc2 data 0xffff [ 596.210773] kvm [2201]: vcpu0, guest rIP: 0xffffffffb885bc98 disabled perfctr wrmsr: 0xc2 data 0xffff Is this a

I have to disagree on that. NATs is the problem and I am one of the causes of that problem as one of the principals behind RFC 1918. What has happened is that HTTP has become the transport for the Internet. Very bad in a number of ways. But for another time. Perhaps. Right now I have to deal with a new ISP that was on the road to static IPv6 when somehow the lead engineer kind of

On 03/06/2015 10:40 AM, Robert Moskowitz wrote: > I have just moved a host from a network that supports static IPv4 and > IPv6. The IPv4 addr is set in ifcfg-eth0, and the IPv6 via RA (I set > the MAC so I get an IPv6 addr that I like). > > I just moved the host to a network that supports static IPv4, but only > dymanic IPv6, so at this time (until I get static IPv6), I need to

On 03/06/2015 11:00 AM, Robert Moskowitz wrote: > > > On 03/06/2015 10:55 AM, Barry Brimer wrote: >> >> >>> IPV6INIT="no" >>> >>> But I am still getting a global IPv6 (and of course local scope). >>> >>> What else do I need to do to disable the listening for RA announcements >>> >>> and setting an IPv6

Try: sysctl -w net.ipv6.conf.all.accept_ra=1 to persist between boots, be sure to add this to your /etc/sysctl.conf file. This should prevent the box from listening to any RA announcements. Chris On Sun, Mar 8, 2015 at 10:55 PM, Ryan Wagoner <rswagoner at gmail.com> wrote: > On Fri, Mar 6, 2015 at 11:52 AM, Robert Moskowitz <rgm at htt-consult.com> > wrote: > > >

No change after running this and trying both: system network restart ifdown eth0; ifup eth0 Still having an IPv6 addr. The box has been up for 140 days. Would like to keep it running... This box is really Redsleeve 6, which is the port of Centos 6 to arm. The kernel I am using is the F19 kernel. All of this MIGHT be contributing to things not working as they would on a 'normal'