于 2011-12-26 23:44, 李晖 写道: > the patch from hanhua did not work for me right now . > But by using 64 bit kernel, I can get the shell, by using logcat, I > found that the dalvik vm did not up, > > D/AndroidRuntime(1315): >>>>>>>>>>>>>> Android Runtime START > D/AndroidRuntime(1315): CheckJNI is ON > E/dalvikvm(1315): ERROR: unexptcted asm interp size 33152 > E/dalvikvm(1315): (did an instruction handler execeed 128 byte? > E/dalvikvm(1315): VM aborting A java bytecode is interpreted with many native instructions, the interpreter use 128 byte for each bytecode to save these native instructions, but in some cases the resulting native codes are more than 32 so they will take more than 128 byte. Just change 128 to 256 or something like that will do. > > did any one came across this problem? > > On 12月24日, 上午12时31分, 李晖 <lihui205...@gmail.com> wrote: >> good, hanhua send me a mail declare that he has made a patch for >> linux-2.6.27.1 to support it for 32 bit loongson, I will pull the >> repository out and patch it, then try it if it works for me. >> Thanks all, I will feed back then. >> >> On 12月23日, 上午10时16分, zhangfx <zhan...@lemote.com> wrote: >> >> >> >> >> >> >> >>> 64bit kernel can support 32 bit userland, just be sure to select o32 >>> support in kernel configuration. Default os uses 64bit kernel and 32bit >>> userland >>> 李晖 <lihui205...@gmail.com>编写： >>> thanks for reply, what I am going to do is patching the linux kernel to fit >>> it to 64bit, what I am wandering is if the kernel is 64bit, should the >>> android file system to be 64bit, or 32bit is ok? On 12月22日, 上午11时26分, >>> zhangfx wrote: > 于 2011-12-22 9:45, 晏华 写道: > > > In 32 bit mode, the pci >>> window of loongson2F is too small for some > > devices such as sis315 > > >>> driver. The sis315 claims 128MB pci memory, while loongson2f has only > > >>> 128M(another 64MB is reserved for legacy use)。 Besides the sis315 > > >>> graphics devies, still some other pci devices need pci resources. If > > >>> you does not carefully to assign these resources. the kernel would > > just >>> report resource allocation failed. > > To resolve this problems in 32bit >>> mode of loongson2f desktop box. > > there are some advices. > > 1. Let PMON >>> do the resource assignment, and kernel just use it. > > 2. use pci high mem >>> (please ref loongson 2f usermanual)just as 64bit > > kenel, and you need to >>> modify kenrel code, use tlb to map sothing. > > I think we should have a >>> working kernel for fuloong that can be compiled > in 32bit? For example, is >>> this one（http://dev.lemote.com/code> /linux_loongson）ok？ Requiring too much >>> knowledge just for a working > kernel is not a good thing. Loongson >>> development should not be so hard > nowadays. > If not, we can just fix it. >>> > > Or the community kernel? Wu zhangjin paid lots of effort to make a > >>> uniform kernel that working for almost all lemote machines(the one > >>> availabe inhttp://dev.lemote.com/trac/linux-loongson-community), not > sure >>> about 32bit configurations(due to the memory limits, it is rarely > used). >>> > > By the way, 64bit kernel can work with android with some patches. > > > >>> > > > > > > > > 2011/12/22 李晖 > > > > Hi, all > > I am porting android to >>> loongson2f, I use "Gcc 4.4.1 + Binutils > > 2.20 on X86" from loongson >>> community compile 2.6.35 linux kernel, by > > setting the kernel model to >>> 64bit, I got the vmlinux image, and it can > > boot up, but some drivers of >>> the kernel are 32bit only(google android > > related), so android can not >>> boot up eventually, so I changed the > > kernel mode to 32bit and did not >>> change other config and rebuild the > > kernel again, the vmlinux can be >>> generated, but when I load and boot > > it by using pmon, the kernel can >>> not boot, the pmon print register > > information and no further >>> information on the screen. > > By adding printk at the begin of >>> start_kernel function, I found it > > even not reached that point! > > > >>> The machine I am using is a loongson2f desktop computer, and I found > > >>> that those who claims had succeeded port android to loongson2f all > > uses >>> loongson2f mini notebook, did it effect? > > > andy tips or suggestion will >>> be greatly appreciated! > > > -- > > You received this message because you >>> are subscribed to the Google > > Groups "loongson-dev" group. > > To post >>> to this group, send email to loongson-dev@googlegroups.com > > . > > To >>> unsubscribe from this group, send email to > > >>> loongson-dev+unsubscr...@googlegroups.com > > . > > For more options, visit >>> this group at > >http://groups.google.com/group/loongson-dev?hl=en. > > > >>> -- > > 晏华 > > 龙芯梦兰 软件部 > > -- > > You received this message because you are >>> subscribed to the Google > > Groups "loongson-dev" group. > > To post to >>> this group, send email to loongson-dev@googlegroups.com. > > To unsubscribe >>> from this group, send email to > > >>> loongson-dev+unsubscr...@googlegroups.com. > > For more options, visit this >>> group at > >http://groups.google.com/group/loongson-dev?hl=en. -- You >>> received this message because you are subscribed to the Google Groups >>> "loongson-dev" group. To post to this group, send email to >>> loongson-dev@googlegroups.com. To unsubscribe from this group, send email >>> to loongson-dev+unsubscr...@googlegroups.com. For more options, visit this >>> group athttp://groups.google.com/group/loongson-dev?hl=en.
-- You received this message because you are subscribed to the Google Groups "loongson-dev" group. To post to this group, send email to loongson-dev@googlegroups.com. To unsubscribe from this group, send email to loongson-dev+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/loongson-dev?hl=en.
