<div dir="ltr"><div dir="ltr">On Mon, Jun 24, 2019 at 3:37 PM Jake Ehrlich <<a href="mailto:jakehehrlich@google.com">jakehehrlich@google.com</a>> wrote:<br></div><div class="gmail_quote"><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div dir="ltr">disclaimer: I work at Google so don't take my +1 as an independent vote forward.<br><br>We would like to use this on Fuchsia and I am particularly interested in creating a dynamic linking library for ELF with Roland McGrath's guidance. We spoke about creating a library for writing dynamic linkers internally and I don't see why this can't be upstreamed.<br></div></blockquote><div><br></div><div>If dynamic linking support is added in a "as a library" fashion, so that it can easily be excluded if not required without affecting the rest of the system, I do not see any problems adding it.</div><div> </div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div dir="ltr">On Fuchsia we critically need support for AArch64; What do you expect to be architecture dependent? I struggled to think of where the architecture and not the operating system was the issue.</div></blockquote><div><br></div><div>I think syscalls are an example of being architecture specific? And, items like program startup and PIE loader are operating system/exe format specific?</div><div><br></div><div>Just for my knowledge, why is answering these questions at a general level important? </div><div> </div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Mon, Jun 24, 2019 at 3:23 PM Siva Chandra via llvm-dev <<a href="mailto:llvm-dev@lists.llvm.org" target="_blank">llvm-dev@lists.llvm.org</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div dir="ltr"><span id="gmail-m_-5430183648132132068gmail-m_8926593389638495630gmail-docs-internal-guid-31026820-7fff-9c8b-7125-459f728330fd"><p dir="ltr" style="line-height:1.38;margin-top:0pt;margin-bottom:0pt"><span style="font-family:Arial;color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">Hello LLVM Developers,</span></p><br><p dir="ltr" style="line-height:1.38;margin-top:0pt;margin-bottom:0pt"><span style="font-family:Arial;color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">Within Google, we have a growing range of needs that existing libc implementations don't quite address. This is pushing us to start working on a new libc implementation.</span></p><br><p dir="ltr" style="line-height:1.38;margin-top:0pt;margin-bottom:0pt"><span style="font-family:Arial;color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">Informal conversations with others within the LLVM community has told us that a libc in LLVM is actually a broader need, and we are increasingly consolidating our toolchains around LLVM. Hence, we wanted to see if the LLVM project would be interested in us developing this upstream as part of the project. </span></p><br><p dir="ltr" style="line-height:1.38;margin-top:0pt;margin-bottom:0pt"><span style="font-family:Arial;color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">To be very clear: we don't expect our needs to exactly match everyone else's -- part of our impetus is to simplify things wherever we can, and that may not quite match what others want in a libc. That said, we do believe that the effort will still be directly beneficial and usable for the broader LLVM community, and may serve as a starting point for others in the community to flesh out an increasingly complete set of libc functionality.</span></p><br><p dir="ltr" style="line-height:1.38;margin-top:0pt;margin-bottom:0pt"><span style="font-family:Arial;color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">We are still in the early stages, but we do have some high-level goals and guiding principles of the initial scope we are interested in pursuing:</span></p><br><ol style="margin-top:0pt;margin-bottom:0pt"><li dir="ltr" style="list-style-type:decimal;font-family:Arial;color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"><p dir="ltr" style="line-height:1.38;margin-top:0pt;margin-bottom:0pt"><span style="background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">The project should mesh with the "as a library" philosophy of the LLVM project: even though "the C Standard Library" is nominally "a library," most implementations are, in practice, quite monolithic.</span></p></li><li dir="ltr" style="list-style-type:decimal;font-family:Arial;color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"><p dir="ltr" style="line-height:1.38;margin-top:0pt;margin-bottom:0pt"><span style="background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">The libc should support static non-PIE and static-PIE linking. This means, providing the CRT (the C runtime) and a PIE loader for static non-PIE and static-PIE linked executables.</span></p></li><li dir="ltr" style="list-style-type:decimal;font-family:Arial;color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"><p dir="ltr" style="line-height:1.38;margin-top:0pt;margin-bottom:0pt"><span style="background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">If there is a specification, we should follow it. The scope that we need includes most of the C Standard Library; POSIX additions; and some necessary, system-specific extensions. This does not mean we should (or can) follow the entire specification -- there will be some parts which simply aren't worth implementing, and some parts which cannot be safely used in modern coding practice.</span></p></li><li dir="ltr" style="list-style-type:decimal;font-family:Arial;color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"><p dir="ltr" style="line-height:1.38;margin-top:0pt;margin-bottom:0pt"><span style="background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">Vendor extensions must be considered very carefully, and only admitted when necessary. Similar to Clang and libc++, it does seem inevitable that we will need to provide some level of compatibility with other vendors' extensions.</span></p></li><li dir="ltr" style="list-style-type:decimal;font-family:Arial;color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"><p dir="ltr" style="line-height:1.38;margin-top:0pt;margin-bottom:0pt"><span style="background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">The project should be an exemplar of developing with LLVM tooling. Two examples are fuzz testing from the start, and sanitizer-supported testing.</span></p></li></ol><br><p dir="ltr" style="line-height:1.38;margin-top:0pt;margin-bottom:0pt"><span style="font-family:Arial;color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">There are also few areas which we do not intend to invest in at this point:</span></p><br><ol style="margin-top:0pt;margin-bottom:0pt"><li dir="ltr" style="list-style-type:decimal;font-family:Arial;color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"><p dir="ltr" style="line-height:1.38;margin-top:0pt;margin-bottom:0pt"><span style="background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">Implement dynamic loading and linking support.</span></p></li><li dir="ltr" style="list-style-type:decimal;font-family:Arial;color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap"><p dir="ltr" style="line-height:1.38;margin-top:0pt;margin-bottom:0pt"><span style="background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">Support for more architectures (we'll start with just x86-64 for simplicity).</span></p></li></ol><br><p dir="ltr" style="line-height:1.38;margin-top:0pt;margin-bottom:0pt"><span style="font-family:Arial;color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">For these areas, the community is of course free to contribute. Our hope is that, preserving the "as a library" design philosophy will make such extensions easy, and allow retaining the simplicity when these features aren't needed.</span></p><br><p dir="ltr" style="line-height:1.38;margin-top:0pt;margin-bottom:0pt"><span style="font-family:Arial;color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">We intend to build the new libc in a gradual manner. To begin with,  the new libc will be a layer sitting between the application and the system libc. Eventually, when the implementation is sufficiently complete, it will be able to replace the system libc at least for some use cases and contexts.</span></p><br><p dir="ltr" style="line-height:1.38;margin-top:0pt;margin-bottom:0pt"><span style="font-family:Arial;color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">So, what do you think about incorporating this new libc under the LLVM project?</span></p><br><p dir="ltr" style="line-height:1.38;margin-top:0pt;margin-bottom:0pt"><span style="font-family:Arial;color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">Thank you,</span></p><p dir="ltr" style="line-height:1.38;margin-top:0pt;margin-bottom:0pt"><span style="font-family:Arial;color:rgb(0,0,0);background-color:transparent;font-variant-numeric:normal;font-variant-east-asian:normal;vertical-align:baseline;white-space:pre-wrap">Siva Chandra and the rest of the Google LLVM contributors</span></p></span><br class="gmail-m_-5430183648132132068gmail-m_8926593389638495630gmail-Apple-interchange-newline"></div>
_______________________________________________<br>
LLVM Developers mailing list<br>
<a href="mailto:llvm-dev@lists.llvm.org" target="_blank">llvm-dev@lists.llvm.org</a><br>
<a href="https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev" rel="noreferrer" target="_blank">https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev</a><br>
</blockquote></div>
</blockquote></div></div>