[Qemu-devel] [PATCH] RTL8139 rx-path fixes
Two small changes: fix the buffer wraparound code for 64k ring-buffers,
and enable (and fix) the crc32 checksum calculation. These changes are
needed for the the Realtek driver that shipped with 32-bit Vista.
Cheers,
Tim.
--
Tim Deegan <[EMAIL PROTECTED]>, XenSource UK Limited
Registered office c/o EC2Y 5EB, UK; company number 05334508
Index: hw/rtl8139.c
===
RCS file: /sources/qemu/qemu/hw/rtl8139.c,v
retrieving revision 1.10
diff -u -r1.10 rtl8139.c
--- hw/rtl8139.c11 Jul 2007 22:48:58 - 1.10
+++ hw/rtl8139.c17 Jul 2007 10:09:03 -
@@ -53,9 +53,8 @@
/* debug RTL8139 card C+ mode only */
//#define DEBUG_RTL8139CP 1
-/* RTL8139 provides frame CRC with received packet, this feature seems to be
- ignored by most drivers, disabled by default */
-//#define RTL8139_CALCULATE_RXCRC 1
+/* Calculate CRCs properly on Rx packets */
+#define RTL8139_CALCULATE_RXCRC 1
/* Uncomment to enable on-board timer interrupts */
//#define RTL8139_ONBOARD_TIMER 1
@@ -747,7 +746,7 @@
int wrapped = MOD2(s->RxBufAddr + size, s->RxBufferSize);
/* write packet data */
-if (wrapped && s->RxBufferSize < 65536 && !rtl8139_RxWrap(s))
+if (wrapped && !(s->RxBufferSize < 65536 && rtl8139_RxWrap(s)))
{
DEBUG_PRINT((">>> RTL8139: rx packet wrapped in buffer at %d\n",
size-wrapped));
@@ -1023,7 +1022,7 @@
/* write checksum */
#if defined (RTL8139_CALCULATE_RXCRC)
-val = cpu_to_le32(crc32(~0, buf, size));
+val = cpu_to_le32(crc32(0, buf, size));
#else
val = 0;
#endif
@@ -1129,7 +1128,7 @@
/* write checksum */
#if defined (RTL8139_CALCULATE_RXCRC)
-val = cpu_to_le32(crc32(~0, buf, size));
+val = cpu_to_le32(crc32(0, buf, size));
#else
val = 0;
#endif
[Qemu-devel] More misspellings
Hi, I spellchecked much of QEMU's source a while ago, and produced a list of misspellings, which I now include below. I could submit a patch to fix these, but I'm not sure what its preferred form would be. Would it be better to submit separate patches for changes to the docs and/or strings, or would one big patch be okay? (Bear in mind I've counted 206 non-distinct misspellings.) I'll understand if this is deemed too large a cosmetic issue to be worth committing anything, but OTOH, it shouldn't be especially disruptive, as any merge conflicts would be trivial to fix. Thanks, -- Stuart Brady accordin, addresss, amounnt, an a, apear, arbitary, arbritration, asynchrnonous, autio, begining, choosen, completly, compliation, confition, decrememnted, defering, dependancy, desciptor, dirst, disconect, enought, enoutered, epilog, exeption, existant, explicitely, feild, fordward, formwer, hadle_packet, heplers, immeriately, inconsistancies, indepedent, inspirated, instruciton, instructon, instrunctions, interrput, interupt_request, intruction, jup, latters, lenght, mecanism, miniscule, minmal, montior, neccessary, neeed, negotion, normaly, occured, occuring, olders, overrite, paramters, parmissible, peripherans, peripherial, positionned, preceeding, precition, prefech, prefered, presrved, previos, priledge, priveladged, priveledged, priveleged, priviledge, privledge, psuedo, puting, regiser, replacd, reservered, ressources, rythm, seconday, shoud, simplicitly, somethin, spliting, stabalized, succesfully, suppored, supresses, supression, teh, tranfer, transmiter, trasfer, unknwon, unkown, unstandard, useable, vairable, weired, whitespaces, writting, yeild
Re: [Qemu-devel] More misspellings
I should send you my grad school papers to spellcheck :) On 7/17/07, Stuart Brady <[EMAIL PROTECTED]> wrote: Hi, I spellchecked much of QEMU's source a while ago, and produced a list of misspellings, which I now include below. I could submit a patch to fix these, but I'm not sure what its preferred form would be. Would it be better to submit separate patches for changes to the docs and/or strings, or would one big patch be okay? (Bear in mind I've counted 206 non-distinct misspellings.) I'll understand if this is deemed too large a cosmetic issue to be worth committing anything, but OTOH, it shouldn't be especially disruptive, as any merge conflicts would be trivial to fix. Thanks, -- Stuart Brady accordin, addresss, amounnt, an a, apear, arbitary, arbritration, asynchrnonous, autio, begining, choosen, completly, compliation, confition, decrememnted, defering, dependancy, desciptor, dirst, disconect, enought, enoutered, epilog, exeption, existant, explicitely, feild, fordward, formwer, hadle_packet, heplers, immeriately, inconsistancies, indepedent, inspirated, instruciton, instructon, instrunctions, interrput, interupt_request, intruction, jup, latters, lenght, mecanism, miniscule, minmal, montior, neccessary, neeed, negotion, normaly, occured, occuring, olders, overrite, paramters, parmissible, peripherans, peripherial, positionned, preceeding, precition, prefech, prefered, presrved, previos, priledge, priveladged, priveledged, priveleged, priviledge, privledge, psuedo, puting, regiser, replacd, reservered, ressources, rythm, seconday, shoud, simplicitly, somethin, spliting, stabalized, succesfully, suppored, supresses, supression, teh, tranfer, transmiter, trasfer, unknwon, unkown, unstandard, useable, vairable, weired, whitespaces, writting, yeild -- -- Jonathan Kalbfeld +1 323 620 6682
[Qemu-devel] [PATCH] fix build error
This appears to be a remnant of when kqemu was non-free. It broke build for me (not sure under which conditions it can be reproduced, but the fix is obvious, see the patch). -- Robert Millan My spam trap is [EMAIL PROTECTED] Note: this address is only intended for spam harvesters. Writing to it will get you added to my black list. diff -ur kqemu-1.3.0~pre11.old/Makefile kqemu-1.3.0~pre11/Makefile --- kqemu-1.3.0~pre11.old/Makefile 2007-07-17 19:37:57.0 +0200 +++ kqemu-1.3.0~pre11/Makefile 2007-07-17 19:36:52.0 +0200 @@ -90,9 +90,9 @@ # called from 2.6 kernel kbuild obj-m:= kqemu.o -kqemu-objs:= kqemu-linux.o kqemu-mod.o +kqemu-objs:= kqemu-linux.o kqemu-mod-$(ARCH).o -$(obj)/kqemu-mod.o: $(src)/kqemu-mod-$(ARCH).o - cp $< $@ +$(obj)/kqemu-mod-$(ARCH).o: + $(MAKE) -C $(obj)/common endif endif # PATCHLEVEL
Re: [Qemu-devel] More misspellings
Stuart Brady wrote: > Hi, > > I spellchecked much of QEMU's source a while ago, and produced a list of > misspellings, which I now include below. > > I could submit a patch to fix these, but I'm not sure what its preferred > form would be. Would it be better to submit separate patches for > changes to the docs and/or strings, or would one big patch be okay? > (Bear in mind I've counted 206 non-distinct misspellings.) Make it a set of patches, split per-file (or per group-of-files if they were tiny otherwise). Thiemo
Re: [Qemu-devel] [PATCH] RTL8139 rx-path fixes
On 7/17/07, Tim Deegan <[EMAIL PROTECTED]> wrote: Two small changes: fix the buffer wraparound code for 64k ring-buffers, and enable (and fix) the crc32 checksum calculation. These changes are needed for the the Realtek driver that shipped with 32-bit Vista. Fix for crc32 first argument may be needed for smc91c111.c and mcf_fec.c as well, if I understand correctly, zlib documentation indeed states that pre- and post-conditioning (one's complement) is performed within crc32 function. -- Kind regards, Igor V. Kovalenko
Re: [Qemu-devel] address translation
Hi Andrzej,
There were actually two methods described in the thread referred to in the
thread to which you were referring in your previous mail. :)
The thread was -
http://thread.gmane.org/gmane.comp.emulators.qemu/16604
I used the patch provided by Stuart Brady (in the thread referred above) -
and I believed that this is the one which helps me extract executed PC
values, until you told me in your previous email that this is _not_ the
case, and they are just the disassembled and not the executed PC trace! Can
you please let me know if you were referring to the same thread or a
different one?
Thanks,
-Shashi.
On 7/13/07, andrzej zaborowski <[EMAIL PROTECTED]> wrote:
On 14/07/07, Shashidhar Mysore <[EMAIL PROTECTED]> wrote:
> Thanks for the reply, Andrzej!
>
> Some clarifications below ...
>
> On 7/13/07, andrzej zaborowski <[EMAIL PROTECTED] > wrote:
> > >
> > >
> > > /* convert one instruction. s->is_jmp is set if the translation must
> > >be stopped. Return the next pc value */
> > > static target_ulong disas_insn(DisasContext *s, target_ulong
pc_start)
> > > {
> > > int b, prefixes, aflag, dflag;
> > > int shift, ot;
> > > int modrm, reg, rm, mod, reg_addr, op, opreg, offset_addr, val;
> > > target_ulong next_eip, tval;
> > > int rex_w, rex_r;
> > >
> > > s->pc = pc_start; // This s->pc is what I am using to extract
all
> > > executed PC values
> >
> > This s->pc doesn't correspond to executed PC values, qemu is a
translator.
>
> Can you tell me what s->pc is, in this context? Is the the start PC of a
It is the PC of the disassembled instruction (hence disasm_insn).
> basic block that is about to be executed? Is there any way in which I
could
> extract the trace of executed PCs in QEMU?
Yes, there was a thread about this about two months ago.
Regards
[Qemu-devel] does anybody have idea about this patch for linux debugging
http://www.h7.dion.ne.jp/~qemu-win/download/qemu-20061108-debug-on-linux.patch/ What's the purpose of this patch? Thanks a lot. best wishes, shizheng
[Qemu-devel] IBM makes AIX 6 beta available for download for everyone
IBM opened up a little bit and makes the beta of AIX 6 available to anyone. Maybe this is an opportunity to get AIX running on QEMU. At the moment QEMU does not yet understand the AIX cd boot format. I tried to boot it using the -kernel option b I tried to boot it in the following ways, but that doesn't work yet (as expected) qemu-system-ppc -m prep -cdrom cd.AIX_6_OpenBeta.0723C1.V1.ISO -boot d sudo mount -o loop cd.AIX_6_OpenBeta.0723C1.V1.ISO /mnt/loop qemu-system-ppc -M prep -kernel /mnt/loop/ppc/chrp/bootfile.exe The -kernel option is meant for linux kernels but I hoped it would take an AIX kernel too. You can download the AIX 6 beta from: http://www-03.ibm.com/servers/aix/6/beta.html (free registration required) > Open beta now available > > The open beta for AIX® 6 is now available. > Go to https://www14.software.ibm.com/iwm/web/cc/earlyprograms/ibm/aix6beta/ > for more information and to download the beta code. > > Overview > > Openness, such as compliance with open standards, has always been an integral > part of the AIX operating system (OS). The next release of AIX, Version 6.1, > extends this openness to the product release process with the first ever AIX > open beta. The open beta will allow a broad set of IBM clients to download > and gain experience with AIX 6 before it becomes generally available. > > An "open" beta for the next AIX release differs from the traditional beta in > three key areas: > > * Almost anyone who is interested will be able to download and install a > pre-release version of AIX 6. By contrast, only a few clients would have the > opportunity to test a new AIX release in a traditional beta. > * Participants in the open beta will not receive traditional support from > IBM. Instead, you access a Web forum to discuss questions and issues. > * The only legal document required for participation in the open beta is > a "click to accept" license agreement that clearly states all program > conditions.
