Brian, Have you tested the latest patch on Qemu?
We should also set BIT10 when disabling all HW breakpoints. Then it could make (NewDr7 == BIT10) happened. My suggestion is as below // DR7 = Disables all HW breakpoints except for DR3 I/O port access of length 1 byte // CR4 = Make sure DE(BIT3) is set // - AsmWriteDr7 (0); + AsmWriteDr7(BIT10); Thanks! Jeff -----Original Message----- From: Brian J. Johnson [mailto:[email protected]] Sent: Thursday, March 19, 2015 12:33 AM To: [email protected] Subject: [edk2] [PATCH v3] PeCoffExtraActionLibDebug: Restore debug registers in PeCoffExtraActionLibDebug PeCoffExtraActionLibDebug uses the debug registers to pass module load information to the DebugAgent, then restores the old register values. However, it was missing code to restore Dr7 in the DEBUG_LOAD_IMAGE_METHOD_SOFT_INT3 case. This broke hardware breakpoints and watchpoints. It could also lose modifications the debugger made to Cr4. Restore the Dr7 and Cr4 values correctly in the DEBUG_LOAD_IMAGE_METHOD_SOFT_INT3 case, as well as the DEBUG_LOAD_IMAGE_METHOD_IO_HW_BREAKPOINT case. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Brian J. Johnson <[email protected]> --- This fix has been built with OvmfPkg/OvmfPkgIa32X64.dsc, using the GCC48 toolchain under Ubuntu. It has been tested (32- and 64-bit mode) using qemu with tcg and a proprietary DebugAgentLib. We've also used it on hardware in a proprietary BIOS with various Intel CPUs, using the SourceLevelDebugPkg DebugAgentLib (an older version). It's been built successfully with the WINDDK compilers, but I'm not really in a position to build with MSVC. A note on Dr7 bit 10: the Intel SDM says that this bit always reads as 1, and that's the behavior I see on hardware. (The latest SDM says it reads as 0 in 32-bit mode, but this appears to be a typo. A case is open with Intel.) However, on simulators, including qemu in tcg mode, it often reads as 0. So we force it to 1 unconditionally to simplify later code. Changes in v2: - Generalized comment wording - Properly restore Cr4 as well Changes in v3: - Set bit 10 when reading Dr7, and simplify the test if Dr7 has changed. This has the side effect of writing Dr7=BIT10 if it was read as 0 (on simulators), but this should be harmless. Thanks, Brian My statements are my own, are not authorized by SGI, and do not necessarily represent SGI’s positions. .../PeCoffExtraActionLib.c | 20 +++++++++++++------- 1 file changed, 13 insertions(+), 7 deletions(-) diff --git a/SourceLevelDebugPkg/Library/PeCoffExtraActionLibDebug/PeCoffExtraActionLib.c b/SourceLevelDebugPkg/Library/PeCoffExtraActionLibDebug/PeCoffExtraActionLib.c index 9bf76bf..f45e7aa 100644 --- a/SourceLevelDebugPkg/Library/PeCoffExtraActionLibDebug/PeCoffExtraActionLib.c +++ b/SourceLevelDebugPkg/Library/PeCoffExtraActionLibDebug/PeCoffExtraActionLib.c @@ -104,7 +104,7 @@ PeCoffLoaderExtraActionCommon ( Dr1 = AsmReadDr1 (); Dr2 = AsmReadDr2 (); Dr3 = AsmReadDr3 (); - Dr7 = AsmReadDr7 (); + Dr7 = AsmReadDr7 () | BIT10; // H/w sets bit 10, some simulators + don't Cr4 = AsmReadCr4 (); // @@ -144,7 +144,7 @@ PeCoffLoaderExtraActionCommon ( // E.g.: User halts the target and sets the HW breakpoint while target is // in the above exception handler // - NewDr7 = AsmReadDr7 (); + NewDr7 = AsmReadDr7 () | BIT10; // H/w sets bit 10, some simulators + don't if (!IsDrxEnabled (0, NewDr7) && (AsmReadDr0 () == 0 || AsmReadDr0 () == Signature)) { // // If user changed Dr3 (by setting HW bp in the above exception handler, @@ -161,11 +161,17 @@ PeCoffLoaderExtraActionCommon ( if (!IsDrxEnabled (3, NewDr7) && (AsmReadDr3 () == IO_PORT_BREAKPOINT_ADDRESS)) { AsmWriteDr3 (Dr3); } - if (AsmReadCr4 () == (Cr4 | BIT3)) { - AsmWriteCr4 (Cr4); - } - if (NewDr7 == 0x20000480) { - AsmWriteDr7 (Dr7); + if (LoadImageMethod == DEBUG_LOAD_IMAGE_METHOD_IO_HW_BREAKPOINT) { + if (AsmReadCr4 () == (Cr4 | BIT3)) { + AsmWriteCr4 (Cr4); + } + if (NewDr7 == 0x20000480) { + AsmWriteDr7 (Dr7); + } + } else if (LoadImageMethod == DEBUG_LOAD_IMAGE_METHOD_SOFT_INT3) { + if (NewDr7 == BIT10) { + AsmWriteDr7 (Dr7); + } } // // Restore original IDT entry for INT1 if it was hooked. ------------------------------------------------------------------------------ Dive into the World of Parallel Programming The Go Parallel Website, sponsored by Intel and developed in partnership with Slashdot Media, is your hub for all things parallel software development, from weekly thought leadership blogs to news, videos, case studies, tutorials and more. Take a look and join the conversation now. http://goparallel.sourceforge.net/ _______________________________________________ edk2-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/edk2-devel ------------------------------------------------------------------------------ Dive into the World of Parallel Programming The Go Parallel Website, sponsored by Intel and developed in partnership with Slashdot Media, is your hub for all things parallel software development, from weekly thought leadership blogs to news, videos, case studies, tutorials and more. Take a look and join the conversation now. http://goparallel.sourceforge.net/ _______________________________________________ edk2-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/edk2-devel
