| Message ID | 201010122134.o9CLYoxu005707@triton8.kn-bremen.de |
|---|---|
| State | New |
| Headers | show |
On 10/12/2010 04:34 PM, Juergen Lock wrote: > In article<4CB38C82.1090403@linux.vnet.ibm.com> you write: > >> After suffering from a prolonged maintainer softlockup, I'm attempting >> to get 0.13.0 release process back on track. >> >> I've tagged qemu-0.13.0-rc3 in git which only carries a few changes >> since 0.13.0-rc1. Most notably, a series of updates from Kevin Wolf and >> Cam Macdonell's ivshmem device. >> >> I think we're pretty good testing wise for the final release but I >> wanted to offer a 24-hour window for last minute fixes. I'm only >> interested in the following: >> >> 1) Patches that are *tested* against the stable-0.13 branch that are >> already committed in master. Please tell me explicitly that you've >> tested the patch and how you've tested it. >> >> [...] >> > First tests: > > - vmmouse seems broken (or disabled? I don't see it in `info qdm'...) > I don't think vmmouse is part of qdev yet. > - to test arm emulation I tried booting my old zaurus images (-M terrier) > which failed (q&d fix below), and in the process also tested -M n800 > w/o an image which failed with: > > qemu-system-arm: Duplicate ID 'null' for chardev > Can't create serial device, empty char device > > and that I got fixed by cherry-picking this commit from master: > > 6a8aabd3c132188ee8e0e82ef4aba09f782cbe96 > > From: Stefan Weil<weil@mail.berlios.de> > Date: Sun, 8 Aug 2010 14:09:26 +0200 > Subject: [PATCH] hw/omap: Fix default setup for OMAP UART devices > Sorry, is the patch below an additional fix? If so, please submit against master. Regards, Anthony Liguori > ========== > > And the zaurus patch, problem was the 2nd scoop's base address > (0x08800040) gets rounded down to start of page which causes its > io read/write callbacks to be passed addresses 0x40 higher than > the code expects: (as witnessed by "Bad register offset" messages > and failure to attach the internal CF disk aka microdrive at least.) > > Signed-off-by: Juergen Lock<nox@jelal.kn-bremen.de> > > --- a/hw/zaurus.c > +++ b/hw/zaurus.c > @@ -70,6 +70,10 @@ static uint32_t scoop_readb(void *opaque > { > ScoopInfo *s = (ScoopInfo *) opaque; > > + // XXX Workaround for base address (0x08800040 in this case) > + // rounded down to start of page > + addr&= 0x3f; > + > switch (addr) { > case SCOOP_MCR: > return s->mcr; > @@ -104,6 +108,10 @@ static void scoop_writeb(void *opaque, t > ScoopInfo *s = (ScoopInfo *) opaque; > value&= 0xffff; > > + // XXX Workaround for base address (0x08800040 in this case) > + // rounded down to start of page > + addr&= 0x3f; > + > switch (addr) { > case SCOOP_MCR: > s->mcr = value; > >
On Tue, Oct 12, 2010 at 05:00:34PM -0500, Anthony Liguori wrote: > On 10/12/2010 04:34 PM, Juergen Lock wrote: > > In article<4CB38C82.1090403@linux.vnet.ibm.com> you write: > > > >> After suffering from a prolonged maintainer softlockup, I'm attempting > >> to get 0.13.0 release process back on track. > >> > >> I've tagged qemu-0.13.0-rc3 in git which only carries a few changes > >> since 0.13.0-rc1. Most notably, a series of updates from Kevin Wolf and > >> Cam Macdonell's ivshmem device. > >> > >> I think we're pretty good testing wise for the final release but I > >> wanted to offer a 24-hour window for last minute fixes. I'm only > >> interested in the following: > >> > >> 1) Patches that are *tested* against the stable-0.13 branch that are > >> already committed in master. Please tell me explicitly that you've > >> tested the patch and how you've tested it. > >> > >> [...] > >> > > First tests: > > > > - vmmouse seems broken (or disabled? I don't see it in `info qdm'...) > > > > I don't think vmmouse is part of qdev yet. > Ok, but it also stopped working, pointer doesn't move... > > - to test arm emulation I tried booting my old zaurus images (-M terrier) > > which failed (q&d fix below), and in the process also tested -M n800 > > w/o an image which failed with: > > > > qemu-system-arm: Duplicate ID 'null' for chardev > > Can't create serial device, empty char device > > > > and that I got fixed by cherry-picking this commit from master: > > > > 6a8aabd3c132188ee8e0e82ef4aba09f782cbe96 > > > > From: Stefan Weil<weil@mail.berlios.de> > > Date: Sun, 8 Aug 2010 14:09:26 +0200 > > Subject: [PATCH] hw/omap: Fix default setup for OMAP UART devices > > > > Sorry, is the patch below an additional fix? Yes, different and unrelated bug. > If so, please submit > against master. > Right, I shall build & test that as well - later. :) Thanx, Juergen
On Wed, Oct 13, 2010 at 05:28:37AM +0200, Juergen Lock wrote: > On Tue, Oct 12, 2010 at 05:00:34PM -0500, Anthony Liguori wrote: > > On 10/12/2010 04:34 PM, Juergen Lock wrote: > > > In article<4CB38C82.1090403@linux.vnet.ibm.com> you write: > > > > > >> After suffering from a prolonged maintainer softlockup, I'm attempting > > >> to get 0.13.0 release process back on track. > > >> > > >> I've tagged qemu-0.13.0-rc3 in git which only carries a few changes > > >> since 0.13.0-rc1. Most notably, a series of updates from Kevin Wolf and > > >> Cam Macdonell's ivshmem device. > > >> > > >> I think we're pretty good testing wise for the final release but I > > >> wanted to offer a 24-hour window for last minute fixes. I'm only > > >> interested in the following: > > >> > > >> 1) Patches that are *tested* against the stable-0.13 branch that are > > >> already committed in master. Please tell me explicitly that you've > > >> tested the patch and how you've tested it. > > >> > > >> [...] > > >> > > > First tests: > > > > > > - vmmouse seems broken (or disabled? I don't see it in `info qdm'...) > > > > > > > I don't think vmmouse is part of qdev yet. > > > Ok, but it also stopped working, pointer doesn't move... > > > > - to test arm emulation I tried booting my old zaurus images (-M terrier) > > > which failed (q&d fix below), and in the process also tested -M n800 > > > w/o an image which failed with: > > > > > > qemu-system-arm: Duplicate ID 'null' for chardev > > > Can't create serial device, empty char device > > > > > > and that I got fixed by cherry-picking this commit from master: > > > > > > 6a8aabd3c132188ee8e0e82ef4aba09f782cbe96 > > > > > > From: Stefan Weil<weil@mail.berlios.de> > > > Date: Sun, 8 Aug 2010 14:09:26 +0200 > > > Subject: [PATCH] hw/omap: Fix default setup for OMAP UART devices > > > > > > > Sorry, is the patch below an additional fix? > > Yes, different and unrelated bug. > > > If so, please submit > > against master. > > > Right, I shall build & test that as well - later. :) Done, and I can report vmmouse is broken on master too. Or at least it is w/ jit, didn't test kvm since this was on FreeBSD. And I tested sidux/aptosid isos since those enable vmmouse by default if detected. Cheers, Juergen
========== And the zaurus patch, problem was the 2nd scoop's base address (0x08800040) gets rounded down to start of page which causes its io read/write callbacks to be passed addresses 0x40 higher than the code expects: (as witnessed by "Bad register offset" messages and failure to attach the internal CF disk aka microdrive at least.) Signed-off-by: Juergen Lock <nox@jelal.kn-bremen.de> --- a/hw/zaurus.c +++ b/hw/zaurus.c @@ -70,6 +70,10 @@ static uint32_t scoop_readb(void *opaque { ScoopInfo *s = (ScoopInfo *) opaque; + // XXX Workaround for base address (0x08800040 in this case) + // rounded down to start of page + addr &= 0x3f; + switch (addr) { case SCOOP_MCR: return s->mcr; @@ -104,6 +108,10 @@ static void scoop_writeb(void *opaque, t ScoopInfo *s = (ScoopInfo *) opaque; value &= 0xffff; + // XXX Workaround for base address (0x08800040 in this case) + // rounded down to start of page + addr &= 0x3f; + switch (addr) { case SCOOP_MCR: s->mcr = value;