<?xml version="1.0"?>
<?xml-stylesheet type="text/xsl" href="/rss.xsl.xml"?>
<rss version="2.0" xmlns:dc="http://purl.org/dc/elements/1.1/">
<channel>
    <title>Changes in hp300</title>
    <link>http://nxr.netbsd.org/rss/src/sys/arch/hp300/</link>
    <description></description>
    <language>en</language>
    <copyright>Copyright 2005</copyright>
    <generator>Java</generator>
    
<item>
    <title>The new pmap_68k is no longer experimental.</title>
    <description></description>
    <pubDate>Sun Aug 16 00:00:00 UTC 2026</pubDate>
    <dc:creator>thorpej</dc:creator>
</item>

<item>
    <title>All M68040 configs require FPSP, and all M68060 configs require M060SP,<br/>so just make M68040 and M68060 dependent on their respective software<br/>packages rather than having to duplicate the options across every kernel<br/>config.</title>
    <description></description>
    <pubDate>Wed Aug 05 01:00:00 UTC 2026</pubDate>
    <dc:creator>thorpej</dc:creator>
</item>

<item>
    <title>Now that all m68k platforms are using the commonn db_memrw.c, declare<br/>that file in the common files.m68k.</title>
    <description></description>
    <pubDate>Mon Jul 06 16:00:00 UTC 2026</pubDate>
    <dc:creator>thorpej</dc:creator>
</item>

<item>
    <title>Provide common defaults for NKMEMPAGES_MIN_DEFAULT and NKMEMPAGES_MAX_DEFAULT<br/>for all m68k platforms.  Defaults for all 68010 come from Sun2, defaults<br/>for everyone else come from hp300, and per-platform overrides for min and<br/>max are provided (and preserved).</title>
    <description></description>
    <pubDate>Fri Jul 03 17:00:00 UTC 2026</pubDate>
    <dc:creator>thorpej</dc:creator>
</item>

<item>
    <title>fix typos in comments.</title>
    <description></description>
    <pubDate>Sat May 23 06:00:00 UTC 2026</pubDate>
    <dc:creator>andvar</dc:creator>
</item>

<item>
    <title>Hide platform vmparam differences away from user-space and modules.<br/>For modules that need to know USRSTACK (exec / compat modules), provide<br/>a way to get that via a variable initialized in cpu_startup().</title>
    <description></description>
    <pubDate>Sun May 03 19:00:00 UTC 2026</pubDate>
    <dc:creator>thorpej</dc:creator>
</item>

<item>
    <title>Centralize the definition of VM_{MIN,MAX}_KERNEL_ADDRESS.</title>
    <description></description>
    <pubDate>Thu Apr 30 05:00:00 UTC 2026</pubDate>
    <dc:creator>thorpej</dc:creator>
</item>

<item>
    <title>Re-factor cpu_reboot().  Platforms can provide optional hooks for halt<br/>(to exit back to firmware) and power-off (to, um, power off).</title>
    <description></description>
    <pubDate>Tue Apr 28 03:00:00 UTC 2026</pubDate>
    <dc:creator>thorpej</dc:creator>
</item>

<item>
    <title>Re-factor all of the badaddr() variants into a single badaddr_read()<br/>that takes an access size and can return a value.  Add a badaddr_write()<br/>that similarly takes a size.</title>
    <description></description>
    <pubDate>Sun Apr 26 00:00:00 UTC 2026</pubDate>
    <dc:creator>thorpej</dc:creator>
</item>

<item>
    <title>Centralize the declaration of "nofault".</title>
    <description></description>
    <pubDate>Sun Apr 26 10:00:00 UTC 2026</pubDate>
    <dc:creator>thorpej</dc:creator>
</item>

<item>
    <title>Re-factor the various m68k platforms' identifycpu() functions, and bring<br/>some consistency to how the CPU/FPU/MMU/cache information is displayed.<br/>Each platform now provides a machine_set_model() function that is called<br/>from cpu_startup_common().  This function is respondible for calling<br/>cpu_setmodel() with platform-specific model description.  Then, common code<br/>is responsible for printing the model information, along with the CPU<br/>and cache info, but separately (and this is no longer included in the<br/>hw.model sysctl node; at least it's now consistent!)<br/><br/>Some examples:<br/><br/>  Qemu 10.1.2 Virt platform<br/>  MC68040+MMU+FPU, 4k+4k on-chip I/D caches<br/><br/>(hw.model == "Qemu 10.1.2 Virt platform")<br/><br/>  Motorola MVME-147<br/>  MC68030 CPU+MMU @ 25MHz, MC68882 FPU<br/><br/>(hw.model == "Motorola MVME-147")<br/><br/>  HP 9000/350<br/>  MC68020 CPU @ 25MHz, HP MMU, MC68881 FPU @ 20MHz<br/>  External 32K virtual-address cache<br/><br/>(hw.model == "HP 9000/350")<br/><br/>  HP 9000/433s<br/>  MC68040 CPU+MMU+FPU @ 33MHz, 4K+4K on-chip I/D caches<br/><br/>(hw.model == "HP 9000/433s")<br/><br/>A hook is provided for customized model banners:<br/><br/>  Model: sun2 {120,170}<br/>  MC68010 CPU @ 10MHz, Sun MMU<br/><br/>(hw.model == "sun2 {120,170}")<br/><br/>  Model: sun3 160<br/>  MC68020 CPU @ 16.67MHz, Sun MMU, MC68881 FPU<br/><br/>(hw.model == "sun3 160")<br/><br/>  Model: sun3 50<br/>  MC68020 CPU @ 16MHz, Sun MMU, emulated FPU<br/><br/>(hw.model == "sun3 50")<br/><br/>An additional hook is provided for more complicated model banners:<br/><br/>  SONY NET WORK STATION, Model NWS-1710, Machine ID #123456<br/>  MC68030 CPU+MMU @ 25MHz, MC68882 FPU<br/><br/>(hw.model == "NWS-1710")<br/><br/>The Sun banners are likely to change once a different way to get the<br/>"kernel_arch" is provided for the installer miniroot script, in which<br/>case they will match more closely what the firmware prints at start-up.<br/><br/>Prep work for adopting the generic banner() routine is included here;<br/>that change will come in a future commit.</title>
    <description></description>
    <pubDate>Thu Apr 23 02:00:00 UTC 2026</pubDate>
    <dc:creator>thorpej</dc:creator>
</item>

<item>
    <title>Re-factor genassym.cf into common m68k definitions and platform-specific<br/>ones.</title>
    <description></description>
    <pubDate>Mon Apr 20 09:00:00 UTC 2026</pubDate>
    <dc:creator>thorpej</dc:creator>
</item>

<item>
    <title>Centralize the declaration of msgbufaddr for m68k platforms.</title>
    <description></description>
    <pubDate>Sun Apr 05 14:00:00 UTC 2026</pubDate>
    <dc:creator>thorpej</dc:creator>
</item>

<item>
    <title>Don't use mmutype as a proxy for cputype when deciding what to do<br/>with the cache.</title>
    <description></description>
    <pubDate>Sun Mar 29 14:00:00 UTC 2026</pubDate>
    <dc:creator>thorpej</dc:creator>
</item>

<item>
    <title>switch_subr.s is now a standlone file, so the boilerplate in each<br/>locore.s that pulled it in is unnecessary.</title>
    <description></description>
    <pubDate>Sun Mar 29 03:00:00 UTC 2026</pubDate>
    <dc:creator>thorpej</dc:creator>
</item>

<item>
    <title>I am giving up attempting to bend the kernel build machinery to my will<br/>so that assym.h can be safely included by libkern assembly sources.  So,<br/>instead, redundantly define __HAVE_M68K_BROKEN_RMC in multiple places so<br/>that all the things that need to key off of it can.</title>
    <description></description>
    <pubDate>Wed Dec 03 02:00:00 UTC 2025</pubDate>
    <dc:creator>thorpej</dc:creator>
</item>

<item>
    <title>Add a trimmed down configuration file for the small 68020-based<br/>machines.</title>
    <description></description>
    <pubDate>Sat Nov 29 23:00:00 UTC 2025</pubDate>
    <dc:creator>thorpej</dc:creator>
</item>

<item>
    <title>Fix various typos, mainly in comments.</title>
    <description></description>
    <pubDate>Sat Nov 29 22:00:00 UTC 2025</pubDate>
    <dc:creator>andvar</dc:creator>
</item>

<item>
    <title>extent(9) -&gt; vmem(9)</title>
    <description></description>
    <pubDate>Sat Nov 29 19:00:00 UTC 2025</pubDate>
    <dc:creator>thorpej</dc:creator>
</item>

<item>
    <title>Remove the advertising clause from Adam Glass's license.<br/><br/>Many thanks to Adam for agreeing to this.</title>
    <description></description>
    <pubDate>Fri Nov 28 14:00:00 UTC 2025</pubDate>
    <dc:creator>nia</dc:creator>
</item>
</channel></rss>

