On Sun, Sep 7, 2014 at 11:17 AM, Jordan Justen <[email protected]> wrote: > On Fri, Sep 5, 2014 at 2:03 AM, Laszlo Ersek <[email protected]> wrote: >> In addition, I could tighten the Length + checksum validation with >> ACPI_BUILD_APPNAME6 and ACPI_BUILD_APPNAME4 checks, according to qemu's >> build_header() function -- if Michael agrees that these are stable. IOW, >> the OEMID would have to be "BOCHS ", and the first four bytes of >> OEMTableID would have to be "BXPC". I think these four checks together >> are pretty strong: a static check for a *10-byte* signature (in effect), >> and a dynamic check for length + checksum. > > Michael, what do you think about Laszlo's idea to verify "BOCHS" and > "BXPC" in the tables? Can we assume that these won't be changing > anytime soon? > > You also suggested a new flag to indicate that a blob is acpi data. I > guess if we ever see that we can skip the extra ACPI table checks, > such as OEMID/OEMTableID.
Laszlo, For your v2 series, Reviewed-by: Jordan Justen <[email protected]> contingent on Michael agreeing that checking OEMID/OEMTableID is okay. Regarding the new flag, it seems like we can add support for that separately if it is implemented. -Jordan ------------------------------------------------------------------------------ Want excitement? Manually upgrade your production database. When you want reliability, choose Perforce. Perforce version control. Predictably reliable. http://pubads.g.doubleclick.net/gampad/clk?id=157508191&iu=/4140/ostg.clktrk _______________________________________________ edk2-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/edk2-devel
